Transcript SB 2.0 poster

Re-engineering transcriptional control in the yeast pheromone
response pathway
Alex
1
Mallet
& Drew
2
Endy
MIT Computational and Systems Biology1 & Biological Engineering2
Abstract
Feedback in the pathway
The pheromone response pathway in S.cerevisiae is regulated on multiple levels and
timescales, and by many biochemical mechanisms. In particular, several proteins that are
directly involved in signal transduction through the pathway, as well as several positive and
negative regulators of the pathway, are transcriptionally upregulated in response to
pheromone.
Opening the feedback loops
Replacement Promoters
STE12-mediated transcriptional feedback
Constitutive promoters
•Library of TEF promoters4
•Allow varying expression level over 2
orders of magnitude
•GPD, TEF, ADH1, BMH2, ACT1, CYC
•Capable of producing ~1000 – 90,000
molecules/cell, depending on protein3
The transcription factor STE12 is activated in response to pheromone and increases transcription
of several proteins involved in the pathway over their basal expression levels. These proteins, in turn,
exert regulatory control on each other, as depicted below. STE12 thus acts as a key component
contributing to a set of interlocking positive and negative feedback loops.
Previous work has focused on the effects of deleting or overexpressing proteins
involved in propagating and modulating the signal. While data from these experiments
sheds some light on the impact of varying protein levels, it does not represent a systematic
exploration of the importance of transcriptional regulation of pathway genes in response
to pheromone. Here, we describe work that explicitly addresses the question whether
specific instances of pheromone-dependent transcriptional feedback loops are necessary
for proper pathway function.
Regulatable promoter
•Combination of artificial promoter based on Tetracycline repressor and artificial transcription factor rtTA(S2) 5
•rtTA(S2) binds DNA in the presence of doxycycline, allowing induction of transcription by addition of
doxycycline
•Allows varying expression level over 3 orders of magnitude, with unimodal cell response and low noise levels6
Determining promoter sequences to replace
STE2
Transcriptional upregulation
Post-translational downregulation
Specifically, we are working to replace the promoters of pheromone-inducible pathway
genes, both singly and in combination, with constitutive and exogenously regulatable
promoters. These custom promoters will allow us to explore pathway response in the
absence of transcriptional feedback, via regulatable expression of some of the major
contributors to signal transmission and control.
GPA1
SST2
Signal transduction
Design rules:
• Get rid of all annotated7 upstream transcription factor binding sites
• Actually remove the sequence, don’t just insert the new promoter
• Get rid of as many potential (i.e. based only on motif match) transcription factor regulation sites as possible
But:
• Avoid disrupting transcription initiation or termination of upstream genes
Case 1: Upstream gene in same orientation as target gene
MSG5
The synthetic pathway will be characterized to determine the effects of these
modifications and to extend our understanding of pathway function. Assaying such
characteristics as mating efficiency, recovery from pheromone-induced cell-cycle arrest and
genome-wide transcription levels will allow us to understand the contributions made by
wild-type transcriptional regulation to pathway function and mating program induction.
Case 2: Upstream gene in opposite orientation as target gene
FUS3
Intergenic region
Intergenic region
 Upstream gene … STOP
STE12
ATG …Target gene 
Upstream gene  GTA
ATG …Target gene 
TF binding sites
 Upstream gene … STOP
FAR1
Pathway overview
1. Over what range of constitutive (i.e. not mediated by STE12) expression of single proteins does the pathway retain proper function ?
2. How robust is the pathway to the removal of multiple transcriptional feedback loops ?
We will answer these questions by removing the STE12-mediated transcriptional feedback loops, both singly and
in combination. Pathway genes that are STE12-responsive in the wild-type pathway will be placed under the control
of promoters that allow us to vary their mRNA levels across a large range, and the response of the altered pathway
will be characterized. The results of these experiments will also allow us to refine our computational model.
Constitutive STE2 synthesis
Pheromone-induced STE2 synthesis
•Adaptation to downregulate pathway activity appropriately
•Resistance to cell cycle arrest when exposed to pheromone outside G1
•Maintenance of specificity e.g. avoiding inappropriate activation of the filamentous growth
pathway
Active GPA1 (molecules/cell)
STE2 (molecules/cell)
4000
3000
600
400
200
0
50
100
150
Time (min)
200
6000
6000
5500
5000
250
0
-50
300
0
50
100
150
Time (min)
200
250
5000
4500
4000
Constitutive STE2 synthesis
Pheromone-induced STE2 synthesis
0
50
The functioning of the re-engineered pathway thus needs to be examined at multiple levels. We will examine pathway function by:
• Monitoring the expression of a chromosomally-integrated copy of YFP driven off the pheromone-responsive PRM1 promoter, which will provide
insight on the timecourse of signaling through the pathway
• Generating dose-response curves of pheromone sensitivity, to characterize the impact on basal signaling
• Examining the efficiency of recovery from cell-cycle arrest upon removal of pheromone, thereby measuring the extent to which pathway
desensitization is affected
• Generating genome-wide transcript measurements, to obtain a comprehensive picture of the molecular response induced by the altered pathway
connectivity
• Assaying mating efficiency, as a means of gauging the overall impact of the changes to the pathway
References and Acknowledgements
300
We would like to thank members of the Endy and Knight labs, Dr. E. Fraenkel and Dr. A. van Oudenaarden at MIT for providing valuable feedback.
Discussions with Dr. F. Winston of Harvard and Drs. K. Benjamin, A. Colman-Lerner, R. Yu and G. Pesce of the Molecular Sciences Institute also
provided key insights. This work was funded by the Computational and Systems Biology Initiative at MIT.
3800
3600
3000
-50
Yeast cells react to pheromone on a wide range of timescales, from a response time on the order of seconds by the signaling cascade component of
the pheromone response pathway, to the actual mating program, which can take several hours to complete. The pheromone response also produces a
wide variety of phenotypes, from morphological changes such as the formation of a mating projection and cell-cycle arrest to the differential
transcription of several hundred genes.
We are interested in understanding the contribution to proper function of the yeast pheromone response pathway made by pheromone-mediated
transcriptional induction of pathway components and regulators. In this poster, we describe our motivating questions and our approach to answering
these questions, namely placing pathway genes under controllable promoters and characterizing the pathway’s response to removal of one or more
transcriptional feedback loops. At present, we have constructed strains containing the STE2 and FAR1 genes under the control of constitutive
promoters and are in the process of characterizing these strains.
1000
3500
Will decide how much to replace on instance-by-instance basis, taking into
account factors like relevance of upstream gene to mating, higher stringency
for putative regulatory sites based purely on motif matching etc
Conclusions
800
100
150
Time (min)
200
250
3400
SST2 (molecules/cell)
•Efficient induction of the cellular program required for mating in response to pheromone
5000
Active FUS3 (molecules/cell)
•A level of basal activity that prevents inappropriate differentiation
6000
2000
Active STE12 (molecules/cell)
Proper control of the pathway thus has several facets:
New promoter ATG …Target gene 
Constitutive STE2 synthesis
Pheromone-induced STE2 synthesis
1000
7000
0
-50
?
1200
8000
Yeast can exist as haploid cells, in two “mating types” termed a and alpha, with mating between
cells of opposite mating types resulting in a single diploid yeast cell. Mating between yeast cells is
mediated by the pheromone response pathway, which is activated when a haploid cell senses the
presence of pheromones emitted by cells of the opposite mating type. Activation of the pathway
results in transient differentiation of the cell via cell cycle arrest in G1, the formation of a mating
projection and differential transcription of several hundred genes1. If mating does not occur, or the
stimulus is removed, the cells re-enter and proceed through the rest of the mitotic cell cycle. In
addition to mediating the response to pheromone, the pathway also shares some molecular
machinery with pathways activated under different conditions, such as the High Osmolarity Glycerol
(HOG) pathway.
Upstream gene  GTA
Characterizing the modified pathway
Previous work has shown that constitutive overexpression of the genes upregulated by STE12 does not, in the
majority of cases, lead to constitutive pathway activation or total loss of signaling upon exposure to pheromone. This
argues that pathway function is robust in the face of above-basal constitutive levels of expression of single genes and
single changes to the transcriptional architecture of the system (i.e. removal of one transcriptional feedback loop). In
addition, a detailed computational model of the pathway (developed in our laboratory2) predicts pathway dynamics
that are largely independent of transcriptional induction of certain genes, as illustrated below for the receptor STE2,
a gene that is upregulated by a factor of five on exposure to pherome1. This naturally leads to two questions:
9000
ATG …Target gene 
A terminator built into the new promoter will cause proper transcription
termination of the upstream gene, so we can replace as much as we
need to of the upstream region.
Are transcriptional feedback loops necessary for
pathway function ?
(K. Benjamin, Molecular Sciences Institute,
unpublished)
New promoter
TF binding sites
4000
3000
2000
3000
(2) T. Thomson, unpublished
2800
(3) K. Benjamin, personal communication.
2600
2400
1000
300
(1) Roberts CJ et al, 2000. Science 287:873.
3200
0
-50
Constitutive STE2 synthesis
Pheromone-induced STE2 synthesis
0
50
100
150
Time (min)
200
250
300
2000
-50
(4) Alper H, Fischer C, Nevoigt E, Stephanopoulos G, 2005. Proc Natl Acad Sci USA 102:12678
Constitutive STE2 synthesis
Pheromone-induced STE2 synthesis
2200
0
50
100
150
Time (min)
200
250
300
(5) Urlinger S et al, 2000. Proc Natl Acad Sci USA 97:7963
(6) Becskei A, Kaufmann BB, van Oudenaarden A, 2005. Nat Genet 435:937
(7) Harbison CT et al, 2004. Nature 431:99