Lecture 17 Gene expression and the transcriptome II

1
Lecture 17Gene expression and the transcriptome
II
2
SAGE

• SAGE Serial Analysis of Gene Expression
• Based on serial sequencing of 15-bp tags that are
  unique to each and every gene
• SAGE is a method to determine absolute abundance
  of every transcript expressed in a population of
  cells

3
SAGE

• 15-bp gene-specific tags are produced by elegant
  series of molecular biology manipulations and
  then concatenated into a single molecule (string)
  for automated sequencing
• By sequencing the concatenated fragments, the
  number of copies of each tag can be counted
• A list of each unique tag and its abundance in
  the population is assembled

4
SAGE

• At least 50,000 tags are required per sample to
  approach saturation, the point where each
  expressed gene (eukaryotic cell) is represented
  at least twice
• SAGE costs about 5000 per sample
• Too expensive to do replicated comparisons like
  is done with microarrays

5
Transcript abundance in typical eukaryotic cell

• lt100 transcripts account for 20 of of total mRNA
  population, each being present in between 100 and
  1000 copies per cell
• These encode ribosomal proteins and other core
  elements of transcription and translation
  machinery, histones and further taxon-specific
  genes
• General, basic and most important cellular
  mechanisms

6
Transcript abundance in typical eukaryotic cell
(2)

• Several hundred intermediate-frequency
  transcripts, each making 10 to 100 copies, make
  up for a further 30 of mRNA
• These code for housekeeping enzymes, cytoskeletal
  components and some unusually abundant cell-type
  specific proteins
• Pretty basic housekeeping things

7
Transcript abundance in typical eukaryotic cell
(3)

• Further 50 of mRNA is made up of tens of
  thousands low-abundance transcripts (lt10), some
  of which may be expressed at less than one copy
  per cell (on average)
• Most of these genes are tissue-specific or
  induced only under particular conditions
• Specific or special purpose products

8
Transcript abundance in typical eukaryotic cell
(4)

• Get some feel for the numbers (can be a factor 2
  off but order of magnitude about right)
• If
• 80 transcripts 400 copies 32,000 (20)
• 600 transcripts 75 copies 45,000 (30)
• 25,000 transcripts 3 copies 75,000 (50)
• Then Total 150,000 mRNA molecules

9
Transcript abundance in typical eukaryotic cell
(5)

• This means that most of the transcripts in a cell
  population contribute less than 0.01 of the
  total mRNA
• Say 1/3 of higher eukaryote genome is expressed
  in given tissue, then about 10,000 different tags
  should be detectable
• Taking into account that half the transcriptome
  is relatively abundant, at least 50,000 different
  tags should be sequenced to approach saturation
  (at least 10 copies per transcript on average)

10
SAGE analysis of yeast (Velculesco et al., 1997)
1.0 0.75 0.5 0.25 0
17 38 45
Fraction of all transcripts
1000 100 10 1
0.1
Number of transcripts per cell
11
SAGE quantitative comparison

• A tag present in 4 copies in one sample of 50,000
  tags, and in 2 copies in another sample, may be
  twofold expressed but is not going to be
  significant
• Even 20 to 10 tags might not be statistically
  significant given the large numbers of
  comparisons
• Often, 10-fold over- or under-expression is taken
  as threshold

12
SAGE quantitative comparison

• A great advantage of SAGE is that method is
  unbiased by experimental conditions
• Direct comparison of data sets is possible
• Data produced by different groups can be pooled
• Web-based tools for performing comparisons of
  samples all over the world exist (e.g. SAGEnet
  and xProfiler)

13
Genome-Wide Cluster AnalysisEisen dataset

• Eisen et al., PNAS 1998
• S. cerevisiae (bakers yeast)
• all genes ( 6200) on a single array
• measured during several processes
• human fibroblasts
• 8600 human transcripts on array
• measured at 12 time points during serum
  stimulation

14
The Eisen Data

• 79 measurements for yeast data
• collected at various time points during
• diauxic shift (shutting down genes for
  metabolizing sugars, activating those for
  metabolizing ethanol)
• mitotic cell division cycle
• sporulation
• temperature shock
• reducing shock

15
The Data

• each measurement represents
• Log(Redi/Greeni)
• where red is the test expression level, and green
  is
• the reference level for gene G in the i th
  experiment
• the expression profile of a gene is the vector
  of
• measurements across all experiments G1 .. Gn

16
The Data

• m genes measured in n experiments
• g1,1 g1,n
• g2,1 . g2,n
• gm,1 . gm,n

Vector for 1 gene
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
Eisen et al. Results

• redundant representations of genes cluster
  together
• but individual genes can be distinguished from
  related genes by subtle differences in expression
• genes of similar function cluster together
• e.g. 126 genes strongly down-regulated in
  response to stress

21
Eisen et al. Results

• 126 genes down-regulated in response to stress
• 112 of the genes encode ribosomal and other
  proteins related to translation
• agrees with previously known result that yeast
  responds to favorable growth conditions by
  increasing the production of ribosomes

22
Partitional Clustering

• divide instances into disjoint clusters
• flat vs. tree structure
• key issues
• how many clusters should there be?
• how should clusters be represented?

23
(No Transcript)
24
Partitional Clustering from aHierarchical
Clustering
we can always generate a partitional clustering
from ahierarchical clustering by cutting the
tree at some level
25
K-Means Clustering

• assume our instances are represented by vectors
  of real values
• put k cluster centers in same space as
  instances
• now iteratively move cluster centers

26
K-Means Clustering

• each iteration involves two steps
• assignment of instances to clusters
• re-computation of the means

27
K-Means Clustering

• in k-means clustering, instances are assigned to
  one and only one cluster
• can do soft k-means clustering via Expectation
  Maximization (EM) algorithm
• each cluster represented by a normal distribution
• E step determine how likely is it that each
  cluster generated each instance
• M step move cluster centers to maximize
  likelihood of instances

28
(No Transcript)
29
Ecogenomics
Algorithm that maps observed clustering behaviour
of sampled gene expression data onto the
clustering behaviour of contaminant labelled gene
expression patterns in the knowledge base
Sample
Compatibility scores


Condition n (contaminant n)
Condition 1 (contaminant 1)
Condition 2 (contaminant 2)
Condition 3 (contaminant 3)
30
Protein-protein interactions
31
Home
back to Chapter 5 page zygote

The Two-Hybrid System for the Detection of
Protein-Protein Interactions
Protein-protein interaction
  If you want to know whether any particular
proteins bound to protein X. Then such proteins
can be found by the yeast two-hybrid system.
The two-hybrid system allows in vivo detection
of protein-protein interactions as well as the
analysis of the affinity of these interactions.
32
Home
back to Chapter 5 page zygote

The Two-Hybrid System for the Detection of
Protein-Protein Interactions
Protein-protein interaction

•  
• Two-hybrid technology exploits the fact that
  transcriptional activators are modular in nature.
  
• Two physically distinct functional domains are
  necessary to get transcription
• a DNA binding domain (DBD) that binds to the DNA
  of the promoter and
• an activation domain (AD) that binds to the
  basal transcription apparatus and activates
  transcription.

33
Home
back to Chapter 5 page zygote

The Two-Hybrid System for the Detection of
Protein-Protein Interactions
Protein-protein interaction

•  
• In the yeast two-hybrid system, the known gene
  encoding X, is cloned into the "bait" vector.
• In this way, the gene for X is placed into a
  plasmid next to the gene encoding a DNA-binding
  domain from some transcription factor.
• For instance, if the gene for X is cloned into
  the pHybLex/Zeo vector, X would be expressed as a
  fusion protein containing bacterially-derived
  LexA DBD.

34
Home
back to Chapter 5 page zygote

The Two-Hybrid System for the Detection of
Protein-Protein Interactions
Protein-protein interaction
  Separately, a second gene (or a library of
cDNAs encoding potential interactors), Y, is
cloned in frame adjacent to an activation domain
of a different transcription factor. For
instance, it could be inserted next to the DNA
encoding the B42 activation domain (AD) in a
"prey" vector such as pYESTrp2.
35
Home
back to Chapter 5 page zygote

The Two-Hybrid System for the Detection of
Protein-Protein Interactions
Protein-protein interaction
  Thus, in one strain of yeast, a known protein X
is fused to the DNA binding domain of a
transcription factor and in another strain,
unknown proteins are fused to the activation
domain of another transcription factor. If one of
the unknown proteins combines with X, it will
bring the AD over to the DBD, and transcription
will be activated. So the plasmids containing the
"bait" (known protein/DBD) and "prey" (unknown
protein/AD) are then placed into a yeast strain,
where a marker gene has a promoter containing the
sequence bound by the bait protein DBD.
36
Home
back to Chapter 5 page zygote

The Two-Hybrid System for the Detection of
Protein-Protein Interactions
Protein-protein interaction
  However, in order to form a working
transcription factor, it needs the AD provided by
the "prey." The only way that it can get this
activation domain is if the known protein X
combines with some unknown protein Y that is
carrying the AD. If the X and Y proteins
interact, the B42 AD is brought into proximity of
the LexA DBD and transcription of the reporter
gene is activated. The activation of the reporter
gene can be screened by enzyme activity, light
release, or cell growth, depending on the type of
reporter gene activated.
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41

• Literature two-hybrid systems
• Fields, S and Song, O. 1989. A novel genetic
  system to detect protein-protein interactions.
  Nature 340245 -246.
• 2. Gyuris, J., Golemis, E., Chertkov, H., and
  Brent, R. 1993. Cdi1, a human G1 and S phase
  protein phosphatase that associates with Cdk2.
  Cell. 75 791-803.