A TWO-COLOUR REPORTER SCREEN AND APPLICATION TO … · A two-colour reporter screen and application...
Transcript of A TWO-COLOUR REPORTER SCREEN AND APPLICATION TO … · A two-colour reporter screen and application...
A TWO-COLOUR REPORTER SCREEN AND APPLICATION TO CELL CYCLE
TRANSCRIPTION
By
Parminder Kainth
A thesis submitted in conformity with the requirements
for the degree of Doctor of Philosophy
Graduate Department of Molecular Genetics
University of Toronto
© Copyright by Parminder Kainth (November 2009)
ii
A two-colour reporter screen and application to cell cycle transcription
Parminder Kainth
Doctor of Philosophy (November 2009)
Department of Molecular Genetics
University of Toronto
Abstract
Development of genome-wide reagents has allowed systematic analysis of gene function.
The experimental accessibility of budding yeast makes it a test-bed for technology
development and application of new functional genomic tools and resources that pave the
way for comparable efforts in higher eukaryotes. In this Thesis, I describe a two-color GFP-
RFP reporter system I developed to assess the consequences of genetic perturbations on a
promoter of interest. The dual-reporter system is compatible with the synthetic genetic array
methodology, an approach that enables marked genetic elements to be introduced into arrays
of yeast mutants via an automated procedure. I use this approach to probe cell cycle-
regulation of histone gene transcription by introducing an HTA1 promoter-GFP reporter gene
construct into an ordered array of ~4500 yeast deletion mutants. I scored defects in reporter
gene expression for each mutant, generating a quantitative analysis of histone promoter
activity. The results of my screen motivated a number of follow-up experiments, including
chromatin immunoprecipitation, transcript profiling and genome-wide analysis of
nucleosome positions, which revealed a previously unappreciated pathway that specifies
regions of repressed chromatin in a cell cycle-sensitive manner. A novel aspect of this
iii
pathway is that it involves histone chaperones and a chromatin boundary element.
Specifically, we discovered that the histone chaperone Rtt106 works with two other
chaperones, Asf1 and the HIR complex, to create a repressive chromatin structure at histone
promoters which is bound by the protein Yta7. It was clear from previous work that Asf1
and HIR repress transcription at HTA1 and that HIR localizes to and functions through a
specific element in histone promoters. However, there was no previous data demonstrating a
role for Rtt106 in cell cycle-dependent gene transcription. In sum, I describe a new genomic
screen that I used to discover a novel pathway regulating cell cycle-dependent transcription.
While I examined histone gene expression as proof-of-principle, my screening system could
be applied to virtually any pathway for which a suitable reporter can be devised. I anticipate
this methodology will enable yeast researchers to collect quantitative data on hundreds of
gene expression pathways.
iv
Acknowledgements
I would like to take this opportunity to thank my supervisor, Dr. Brenda Andrews, for
all of her help and support with all aspects of my graduate experience. Without Brenda’s
guidance and patience, this work would not be possible. In addition to guiding me daily with
my project, Brenda gave me the opportunity to work with and meet other researchers, which
has enriched my learning experience over the past four years. I also benefitted from advice
and guidance provided by my committee members Dr. Timothy Hughes and Dr. Frank
Sicheri. I have been lucky to collaborate daily with Tim and members of his laboratory
which has been indispensible for progress with my work.
I spent the first two years of graduate school working with Holly Sassi, a Research
Associate in our laboratory who provided help and guidance during the early years of
graduate school. I would also like to acknowledge Dr. Jeffrey Fillingham, a former
postdoctoral fellow in Dr. Jack Greenblatt’s laboratory. Working with Jeff over the past two
years has been a great experience and all of his advice and instruction will not be forgotten.
The support from friends I have made in the program has enriched my experience in
the PhD program and their help along the way is greatly appreciated. Finally, I must
acknowledge the support from my family throughout my entire education.
v
Table of Contents
Abstract……………………………………………………………………….ii-iii Acknowledgements…………………………………………………………...iv List of Tables………………………………………………………………….viii List of Figures…………………………………………………………………ix-x List of Abbreviations………………………………………………………….xi-xiii Chapter 1: Introduction……………………………………………………..1-46 1.1 Gene expression microarrays……………………………………...2-8 1.2 Chromatin immunoprecipitation followed by microarray
hybridization (ChIP-chip)………………………………………….8-16
1.3 Protein binding microarrays………………………………………..17-20 1.4 Genome-wide nucleosome occupancy……………………………..20-26 1.5 The yeast deletion array and the synthetic genetic array
(SGA) approach…………………………………………………….27-33 1.6 Reporter-based screens……………………………………………..33-39 1.7 Global cell cycle transcription……………………………………...39-45 1.8 Summary and overall significance………………………………….45-46 Chapter 2: Comprehensive genetic analysis of transcription factor pathways using a dual reporter gene system in budding yeast……………..47-69 Abstract………………………………………………………………………….48 2.1 Introduction…………………………………………………………49-52 2.2 Description of methods……………………………………………..52-60
2.2.1 Reporter system…………………………………………...52-53
vi
2.2.2 Generating an output array of GFP and RFP reporter genes in yeast deletion mutants………………………….54-57
2.2.3 Digital imaging of yeast plates…………………………..57
2.2.4 Assaying GFP and RFP fluorescence intensities from colonies arrayed on agar plates………………………….58-59 2.2.5 Quantifying fluorescence intensities from output scans….59-60
2.3 Analysis of data from a genome-wide promoter-reporter screen…..60-69
2.3.1 Using colony size data to filter dead/sick colonies from further analysis………………………………………62-63 2.3.2 Normalization of GFP and RFP intensities……………….64
2.3.3 Correlation between replicate screens and display of genome-wide data………………………………………….64-69
2.4 Concluding remarks………………………………………………….69
Chapter 3: A two-colour cell array screen reveals interdependent roles for histone chaperones and a chromatin boundary regulator in histone gene repression………………………………………………………70-114 Abstract…………………………………………………………………………...71
3.1 Introduction…………………………………………………………..72-76
3.2 Experimental Procedures……………………………………………..76-80
3.2.1 Yeast strains and plasmids………………………………….76-77
3.2.2 SGA-based Functional Genomic Screen for Regulators of HTA1 expression……………………………..78 3.2.3 qPCR Analysis of Histone Gene Expression………………..78-79
3.2.4 Chromatin Immunoprecipitation (ChIP)……………………79
3.2.5 Purification and Analysis of Rtt106-associated proteins…...80
3.2.6 Genome-wide nucleosome occupancy……………………….80
vii
3.3 Results………………………………………………………………….80-106
3.3.1 A dual-reporter functional genomic screen to discover new regulators of gene expression…………...………………80-81
3.3.2 Identification of regulators of HTA1 expression……………..81-87 3.3.3 Rtt106 represses HIR-regulated histone genes……………....88-93
3.3.4 The HTA1-HTB1 promoter region is nucleosome-free in asf1, hir1, rtt106 mutants……………………….........94-98 3.3.5 HIR/RTT106 repression at HTA1-HTB1 creates a requirement for RTT109……………………………………98-100 3.3.6 Yta7 is a boundary element within the HTA1-HTB1 locus......101-106
3.4 Discussion………………………………………………………………107-114
Summary and Future Directions………………………………………………….115-130 4.1 Summary………………………………………………………………..116-119
4.2 Future directions………………………………………………………...119-130
4.2.1 Characterizing the Rtt106/Asf1/HIR/Yta7 pathway genome-wide…………………………………………………..119-121
4.2.2 Characterizing protein domains in Rtt106 required for function with HIR……………………………………….....121-122
4.2.3 Screening overexpression arrays……………………………...123-125 4.2.4 Increasing throughput of reporter-gene analysis using pooled screens in yeast and higher eukaryotes………………..125-129
4.3 Overall significance……………………………………………………...129-130
References……………………………………………………………………………131-152
viii
List of Tables
Table 1-1: Key regulators of cell cycle transcription. Table 2-1: Query strains and plasmids used in the two-colour promoter-reporter screening
system. Table 3-1: Strains used in this Chapter.
ix
List of Figures
Figure 1-1: A two-colour microarray experiment to identify differentially expressed genes.
Figure 1-2: Schematic of the chromatin immunoprecipitation (ChIP) approach.
Figure 1-3: Regulator-gene motifs identified from ChIP-chip analysis of transcription factors.
Figure 1-4: Schematic of a protein binding microarray experiment.
Figure 1-5: Schematic representation of genome-wide nucleosome occupancy experiment.
Figure 1-6: Gene-deletion strategy for replacing each yeast ORF with a kanamycin resistant cassette (KanMX).
Figure 1-7: The synthetic genetic array approach used for high-throughput double mutant
strain construction. Figure 1-8: A forward genetic reporter screen to identify regulators of a promoter of
interest. Figure 2-1: Overview of the dual reporter SGA methodology.
Figure 2-2: Representative fluorescence scan of a single output array plate.
Figure 2-3: Colony size distribution of yeast deletion mutants.
Figure 2-4: Pearson correlation between replicate CLN2pr-GFP screens.
Figure 2-5: Screening deletion mutants to identify regulators of the CLN2 promoter.
Figure 3-1: Reporter-Synthetic Genetic Array (R-SGA) functional genomic screen for regulators of HTA1 expression.
Figure 3-2: Rtt106, Rtt109 and Yta7 regulate histone gene expression.
Figure 3-3: Rtt106 and HIR localize to the promoter region of HTA1-HTB1.
Figure 3-4: HIR and Asf1 are required for Rtt106 localization to HTA1-HTB1.
Figure 3-5: Asf1, HIR and Rtt106 collaborate to assemble chromatin at the HTA1-HTB1 promoter.
Figure 3-6: Constitutive repression at HTA1-HTB1 creates a requirement for RTT109.
x
Figure 3-7: Spt4, Spt5, Spt6 and FACT cross-link to the transcribed regions of HTA1 but not to the promoter region.
Figure 3-8: Yta7 localizes to the HTA1-HTB1 locus.
Figure 3-9: Yta7 creates a boundary within the HTA1-HTB1 locus.
Figure 3-10: A model describing histone chaperone mediated repression at the HTA1 locus in yeast.
Figure 4-1: Overexpression screen to identify regulators of a promoter of interest.
Figure 4-2: Reporter screening using barcoded gene disruption libraries and FACS.
xi
List of Abbreviations
α……………………….alpha
Δ…………………….....gene deletion
µg……………………...microgram
~……………………….approximately
Arg……………………..arginine
CDK……………………cyclin-dependent kinase
cDNA…………………..complementary DNA
ChIP…………………….chromatin immunoprecipitation
ChIP-chip……………….chromatin immunoprecipitation followed by microarray hybridization
ChIP-seq………………...chromatin immunoprecipitation followed by sequencing DaMP…………………...decreased abundance of messenger RNA by perturbation
DN……………………....down
DNA…………………….deoxyribonucleic acid
dSLAM………………….diploid-based synthetic lethality analysis on microarrays
FACS……………………fluorescence activated cell sorting
FAIRE…………………..formaldehyde assisted isolation of regulatory elements
GFP……………………..green fluorescent protein
H3 K56Ac………………histone H3 lysine 56 acetylation
H3 K9Ac………………..histone H3 lysine 9 acetylation
HAT…………………….histone acetyltransferase
HDAC…………………..histone deacetylase
HphMX…………………hygromycin B resistance cassette
xii
HU………………………hydroxyurea
His………………………histidine
KanMX……………........kanamycin resistance cassette
Leu………………………leucine
Lys………………………lysine
MBF…………………….Mlu1 cell cycle box binding factor
mRNA………………….messenger RNA
NAT…………………….nourseothricin
NDR…………………….nucleosome-depleted region
NEG…………………….negative
ORF…………………….open reading frame
PAC…………………….polymerase A and C
PBM……………………protein binding microarray
PCR…………………….polymerase chain reaction
PMT…………………….photomultiplier tube
Pol II……………………RNA polymerase II
pr………………………..promoter
qPCR……………………quantitative polymerase chain reaction
RFP……………………..red fluorescent protein
RNA…………………….ribonucleic acid
RNAi……………………ribonucleic acid interference
R-SGA…………………..reporter-synthetic genetic array
SBF……………………...SCB binding factor
SCB……………………...Swi4,6-dependent cell cycle box
SDS-PAGE………………sodium dodecyl sulphate polyacrylamide gel electrophoresis
xiii
SGA………………….......synthetic genetic array
STAGE…………………...sequence tag analysis of genomic enrichment
TAP……………………….tandem affinity purification
TBP……………………….TATA binding protein
tdTomato………………….tandem dimer tomato
TSS………………………..transcriptional start site
UAS……………………....upstream activating sequence
URS……………………….upstream regulatory sequence
UTR……………………….untranslated region
WT…………………….......wild type
1
Chapter 1
Introduction
2
Proper control of gene expression is vital for virtually all aspects of cellular function.
In many cases, aberrant gene expression is apparent in various diseased states, including
many types of cancers (Weeraratna, 2005). Exploration of mechanisms of transcriptional
control in Saccharomyces cerevisiae has led to major discoveries on how gene expression is
regulated, which are generally applicable in higher organisms because of the high degree of
conservation of the transcriptional machinery (Woychik and Hampsey, 2002). Furthermore,
the relatively straightforward genetics that can be employed in yeast make it an excellent
model organism for testing and validating new technologies (Bader et al., 2003).
The development of functional genomic tools over the last decade or so has made
large-scale systematic analysis of gene function possible. Perhaps one of the most
groundbreaking advances was the development of DNA microarrays, which revolutionized
the gene expression field. Below I summarize applications and discoveries pertinent to gene
regulation that stem from DNA microarray technology, mainly in terms of key studies
carried out in yeast. I then discuss gene expression reporter screens and their application to
cell cycle transcription, a theme that forms the basis of the work carried out in this Thesis.
1.1 Gene expression microarrays
DNA chip technology has enabled transcriptional profiling of the entire complement
of cellular mRNA in parallel. The DNA probes on the microarray are covalently attached to
the support material, and can include DNA fragments (e.g cDNA microarrays) or
oligonucleotides, which can be spotted onto specific sites on the array or synthesized in situ
(Hughes et al., 2001; Ramsay, 1998 and reviewed in Hughes and Shoemaker, 2001). To
interrogate the array, sample RNA is prepared, reverse transcribed to cDNA with fluorescent
3
labels and hybridized to the microarray. Because DNA probes are arranged in known
positions on the microarray, fluorescent signal captured from hybridization of labelled
sample cDNA is indicative of the presence of that particular transcript (Schena et al., 1995;
Shalon et al., 1996). An important attribute of DNA chips is that multiple fluorescent-
labelled samples can be simultaneously hybridized to the array (Schena et al., 1995; Shalon
et al., 1996). This allows, for example, differently labelled samples from a mutant strain and
a wild type strain to be hybridized to the same gene chip (Figure 1-1). In this type of two-
colour experiment, the ratio of signal intensity from the mutant strain compared to the wild
type strain can be directly compared to determine genes that are differentially expressed in a
particular mutant.
4
Figure 1-1: A two-colour microarray experiment to identify differentially expressed genes. RNA is prepared from a wild type strain and a strain deleted for a particular gene of interest. The RNA is reverse transcribed to cDNA and labelled with different fluorescent dyes. The resulting cDNA from both wild type and mutant strains are hybridized to a microarray that contains oligonucleotide sequences complementary to all genes. After scanning fluorescence intensities, differentially expressed genes are identified. Yellow spots indicate equal expression of that particular gene in the mutant and wild type strain. Green indicates lower transcript levels of that gene in the mutant strain compared to wild type while red indicates higher expression of that particular gene in the mutant compared to wild type.
5
Spotted cDNA or oligonucleotide arrays are advantageous because they can be
printed in-house and customized arrays can be produced that are tailored to the particular
experiment being carried out. However, these types of spotted arrays are not printed with the
same probe density as commercial arrays, which are generally made by synthesizing short
oligonucleotide probes directly on the array surface (see Lipshutz et al., 1999 for a review on
manufacturing high density arrays). In addition, arrays synthesized in situ have been shown
to generate the most reproducible results between different laboratories (Bammler et al.,
2005). Commercial arrays are available from various companies including Affymetrix and
Agilent Technologies. The Affymetrix platform requires hybridization of the labelled
experimental and control sample to different microarrays while the Agilent arrays allow two-
colour hybridization on the same array similar to the scheme shown in Figure 1-1. In tests to
determine reproducibility of microarray results between various technologies, single-colour
platforms produced data with greatest precision and Affymetrix arrays in particular were
slightly more sensitive in detecting small changes in gene expression (de Reynies et al.,
2006). However, Agilent arrays have the advantage that half as many microarrays are
required for an experiment since test and control samples are hybridized to the same array.
In pioneering work, Schena et al. (1995) showed the utility of a two-colour cDNA
microarray experiment to identify differentially expressed genes in Arabidopsis thaliana.
They showed that small amounts of initial RNA (2 µg) were required to identify 45
differentially expressed transcripts and also demonstrated the specificity of the array
hybridization by including yeast TRP4 and rat glucocorticoid receptor cDNA, which showed
no hybridization signal (Schena et al., 1995). Later, genome-scale microarrays were
produced for yeast by spotting PCR products corresponding to each ORF on glass slides and
6
allowed characterization of gene expression programs in different conditions (Lashkari et al.,
1997).
Since this initial work, genome-wide expression profiling has become a common
practice in the study of gene regulation. Numerous publications catalogue genome-wide
transcriptional responses to mutations, environmental changes and drug treatment across the
spectrum of model organisms. These types of experiments generate enormous data sets that
are rich in biological information. An important publication by Hughes et al. (2000a)
highlighted the use of genome-wide expression profiling as a phenotypic readout that can be
used to assign gene function. They assembled a compendium of 300 yeast expression
profiles derived from either treatment of cells with chemicals or from assaying an array of
strains with mutations in known and uncharacterized genes (Hughes et al., 2000a). This
reference compendium of expression profiles was used to assign functions to genes based on
similarity of transcriptional responses which was determined by cluster analysis. For
example, strains with gene mutations in components of the ergosterol biosynthesis pathway
(erg2Δ, erg3Δ and tetracycline repressible-ERG11) had similar transcriptional profiles
causing the mutant strains to cluster together in the reference set, consistent with the known
role for the ERG genes in the same cellular process. Deleting the uncharacterized ORF
YER044c resulted in a transcriptional output similar to the ERG mutants, indicating that
YER044c might also participate in ergosterol biosynthesis (Hughes et al., 2000a). Follow-up
experiments revealed that YER044c (now named ERG28) mutants were defective in
ergosterol accumulation compared to wild type cells, supporting the hypothesis derived from
comparative transcriptional profiling (Hughes et al., 2000a).
7
The compendium approach was also successful in identifying drug targets (Hughes et
al., 2000a). In principle, deleting a drug target should have the same phenotypic
consequence as inhibition of the target by drug treatment. Also, drugs that affect similar
cellular processes should result in similar expression profiles and thus co-cluster in a
compendium matrix. In the compendium of expression profiles, clustering was observed
among the erg2Δ mutant and cells treated with dyclonine, an anaesthetic commonly used in
throat lozenges. Dyclonine-treated cells accumulated fecosterol, indicating Erg2 function
was comprised in the presence of dyclonine, thus identifying Erg2 as the target of dyclonine
(Hughes et al., 2000a). These findings show the power of gene expression profiling for
functional discovery and highlight the utility of the compendium approach for studying gene
function.
Gene expression microarrays have also been used in clinical applications, particularly
in the study of cancers. In one study, expression profiles were generated for 13 dissected
human breast tumours (Perou et al., 1999). Variation in gene expression programs among
these different samples, compared to human mammary epithelial cells, revealed that different
breast tumours could be classified and identified based on the observed expression profiles
(Perou et al., 1999). This type of gene expression profiling has been carried out in hundreds
of different tumours and mining these data has been used to generate hypotheses about
cancer biology. For example, computational analysis of hundreds of tumour gene expression
profiles revealed a role for the transcription factor C/EBPβ in regulating genes with aberrant
expression profiles upon cyclin D1 overexpression, a common signature of many tumours
(Lamb et al., 2003).
8
A general concern with data obtained from expression profiling when testing for
regulators is whether or not target genes are directly controlled by the regulator or if the
observed expression response is indirect. For example, suppose a gene expression
microarray carried out on a strain deleted for gene A reveals altered expression of gene X.
The microarray result alone cannot distinguish between direct regulation of gene X by
binding of gene product A to its promoter or an indirect mechanism that involves gene
product A binding to the promoter of another gene whose product directly regulates gene X.
To differentiate direct versus indirect regulation, it is useful to examine the location of
regulatory proteins on DNA. In the next section, I summarize assays for genome-wide
assessment of the localization of regulatory proteins.
1.2 Chromatin immunoprecipitation followed by microarray hybridization (ChIP-
chip)
Control of gene expression requires proper localization of trans-acting regulatory
factors to their cognate cis-elements in promoters of genes. Thus, an important question in
gene regulation is: what regions of DNA are occupied by transcription factors and other
proteins that regulate transcription? A biochemical approach called chromatin
immunoprecipitation (ChIP) has been applied productively to address this question (Orlando,
2000). This approach relies on immunoprecipitation of a tagged protein of interest along
with its associated proteins and DNA (Figure 1-2). Briefly, tagged proteins are cross-linked
to interacting proteins and DNA in vivo by treatment with formaldehyde, chromatin is
prepared, sheared by sonication, protein-protein and protein-DNA complexes are
immunoprecipitated, cross-linking is reversed, DNA is purified and potential target genes of
9
interest are examined by PCR amplification (Orlando, 2000). Shearing of DNA to usually
less than 500 base pairs followed by immunoprecipitation with an antibody directed towards
the tagged protein allows enrichment of only those genomic regions occupied by the protein.
The advantage of this approach is that tagged proteins do not necessarily need to contact the
DNA but can be bridged by other proteins.
10
Figure 1-2: Schematic of the chromatin immunoprecipitation (ChIP) approach. A tagged ORF is expressed and formaldehyde treatment cross-links protein-DNA interactions. Chromatin is prepared, sheared and the tagged protein is immunoprecipitated using an antibody that recognizes the protein tag. In parallel a control input DNA sample is prepared. The cross-link is reversed, DNA is purified and the population of DNA immunoprecipitated with the protein of interest is characterized by PCR analysis of specific regions of the DNA or the entire complement of immunoprecipitated DNA could be determined by comparing the microarray hybridization signal to the control input DNA signal (ChIP-chip).
11
In early work, formaldehyde-mediated cross-linking of proteins to DNA followed by
immunoprecipitation led to insights into the chromatin structure of highly transcribed genes.
Studies in Drosophila melanogaster revealed that perturbed chromatin structure underlies
heat shock-mediated induction of the HSP70 gene (Solomon et al., 1988). Upon heat shock,
fewer proteins were found to cross-link to the transcribed region. This was consistent with
the notion that actively transcribed genes are devoid of histones, likely to allow RNA
polymerase passage and transcriptional elongation. However, the rate of transcription did
not influence histone H4 cross-linking efficiency to the region providing evidence that not all
transcribed genes are depleted of histones, which contested the general view at the time
(Solomon et al., 1988).
The development of DNA chip technology (discussed above), combined with the
ChIP approach, provides the tools to carry out genome-wide protein-promoter localization
studies. In early experiments, DNA chips were produced with probes complementary to
known intergenic or promoter regions, in contrast to the cDNA chips used to characterize the
complement of mRNA transcripts. The immunoprecipitated DNA from a ChIP experiment
is then hybridized to intergenic DNA chips to identify all of the promoter regions a particular
protein cross-links to genome-wide, an approach known as ChIP-chip (Figure 1-2). ChIP-
chip experiments were reported in Saccharomyces cerevisiae, whereby the genomic
localization of the DNA binding proteins Gal4 and Ste12 (Ren et al., 2000), as well as MBF
and SBF (Iyer et al., 2001) (see Section 1.7) was determined. These studies described the
ChIP-chip methodology and showed that target genes of DNA-associated proteins could be
identified. The Gal4 transcriptional activator, which activates genes involved in galactose
metabolism when cells are grown in galactose, bound 10 genes (Ren et al., 2000). Seven of
12
these genes were previously characterized targets of Gal4 while 3 were new (Ren et al.,
2000). The localization of Ste12, a protein that is required for induction of >200 genes upon
activation of the pheromone response pathway, revealed that 29 of these genes are bound by
Ste12 (Ren et al., 2000). In future work, Ste12 localization was revisited using advanced
sequencing technologies or microarrays and identified hundreds of Ste12 binding sites
(Borneman et al., 2007; Harbison et al., 2004; Lefrancois et al., 2009) that were altered in
different conditions tested in a study by Harbison et al. (2004).
In the past 10 years or so, a huge effort has been made to characterize the
transcriptional regulatory network in yeast by charting genome-wide localization of all
known DNA binding proteins using the ChIP-chip methodology (Harbison et al., 2004; Lee
et al., 2002). In the first study, genes encoding 106 yeast transcription factors were tagged
with the myc epitope to create a library of strains suitable for ChIP experiments aimed at
identifying their localization on promoters genome-wide (Lee et al., 2002). Approximately
4000 regulator-promoter interactions were discovered (P<0.001), where 37% of yeast genes
bound at least one of the 106 regulators tested (Lee et al., 2002). A general feature of yeast
promoters discovered here is that many promoters bound multiple regulators suggesting that,
as in higher eukaryotes, combinations of transcription factors control gene expression.
Dissecting data produced from this study revealed six network motifs in yeast cells that
describe general principles of regulation that underlie the transcriptional regulatory network
(Figure 1-3). These are: (1) autoregulation where the regulator controls its own promoter;
(2) multi-component loop where the regulator controls a gene and that gene product controls
the promoter of the initial regulator; (3) feedforward loop where a regulator activates another
regulator and both of these regulators control a common downstream gene; (4) single input
13
motif where a regulator controls expression of a set of genes whose protein products
coordinate the same cellular process; (5) multi-input motif where a group of regulators bind
together to a group of promoters and; (6) regulator chain consisting of a regulator that
activates another regulator which in turn regulates another gene (Lee et al., 2002).
14
Figure 1-3: Regulator-gene motifs identified from ChIP-chip analysis of transcription factors. Circles represent proteins while rectangles represent genes. See text for details. Figure adapted from Lee et al., 2002.
15
In 2004, another study was published where 204 tagged transcription factors were
analyzed for their genome-wide localization (Harbison et al., 2004). ChIP of all 204
transcription factors was performed in rich media, but because environmental conditions
influence gene expression programs, 84 regulators were also tested in a minimum of 1 out of
12 other conditions (Harbison et al., 2004). This project produced an enormous data set
cataloguing 11 000 regulator-promoter interactions which led to the discovery of sequence
motifs in promoters of target genes for 116 regulators. Testing localization of regulators in
different conditions revealed important behavioural patterns of regulator-gene interactions
that are applied broadly across the genome. These patterns are: (1) condition invariant, in
which the regulator binds the same promoters in two different conditions; (2) condition
enabled, in which the regulator does not bind promoters in one condition but in a second
condition the regulator binds many promoters; (3) condition expanded, in which the regulator
binds a set of promoters in one environment and in another environment the regulator binds
that same set and additional promoters; and (4) condition altered, in which the regulator has
preference for a different set of promoters in two different conditions (Harbison et al., 2004).
Histone modifications such as acetylation and methylation of specific lysine residues
also influence transcriptional output. Antibodies with specificity for these modifications
have allowed ChIP-chip experiments to be carried out to determine the regions of DNA
where these various modifications exist (Pokholok et al., 2005). For example, antibodies that
specifically recognize histone H3 lysine 9 and lysine 14 acetylation were used in ChIP-chip
experiments which revealed that these chromatin marks are found particularly at
transcriptional start sites of active genes and the presence of these acetyl lysines correlates
with transcriptional output (Pokholok et al., 2005). In human cells, antibodies specific for
16
methylation marks on histone H3 lysine 4 revealed that enhancer regions are generally
mono-methylated while promoter regions are tri-methylated, providing a powerful means to
identify regulatory regions in more complex genomes (Heintzman et al., 2007).
Technological advances have allowed production of extremely dense DNA
microarrays that allow analysis of whole genomes such as yeast with high resolution (see
David et al., 2006 and Section 1.4). For larger genomes, approaches that utilize sequencing
to characterize ChIP-enriched DNA were developed. For example, an approach called
sequence tag analysis of genomic enrichment (STAGE) was applied to characterize locations
of regulatory proteins along the genome (Kim et al., 2005). Briefly, this approach allowed a
library to be created where each ChIP-enriched DNA molecule is ligated to a sequencing tag.
Many of these molecules are then ligated to form concatemers of ChIP DNA that can be
sequenced in order to identify genomic regions bound by regulatory proteins (Kim et al.,
2005).
Cost-effective next generation sequencing platforms like the Illumina Genome
Analyzer II are becoming the method of choice for characterizing immunoprecipitated DNA
from ChIP experiments, a method referred to as ChIP-seq. ChIP-seq experiments allow
greater resolution of protein localization on DNA, sequencing can be carried out on a small
amount of ChIP DNA and this approach requires less handling than ChIP-chip.
Additionally, for small genomes like yeast, different experiments can be multiplexed into a
single sequencing run for cost effective generation of high quality data (Lefrancois et al.,
2009). Applications of ChIP-seq in the context of work carried out in this Thesis are
discussed further in the Future Directions section of Chapter 4.
17
1.3 Protein binding microarrays
A key to understanding gene regulation is determination of the cis-regulatory
sequences recognized by DNA binding transcription factors. One method to determine
binding sites of transcription factors in promoters is to search for conserved sequence motifs
present in promoters that are occupied by regulatory proteins in ChIP-chip experiments
(Harbison et al., 2004) or to search for common sequence elements in promoters of genes
that share similar expression profiles (Spellman et al., 1998). As noted above, in yeast, most
known transcription factors have been tested for their genome-wide localization using ChiP-
chip under a variety of conditions (Harbison et al., 2004). However, this in vivo approach
has not created a complete map of sequence motifs bound by transcription factors, likely
because transcription factor binding is not occurring under the conditions tested.
Additionally, in ChIP experiments, proteins do not necessarily need to contact DNA but their
association with DNA can be bridged by other regulatory proteins.
Proteome-level approaches have been developed and applied productively to identify
consensus sites that are bound by regulatory proteins. Protein binding microarray (PBM)
experiments have been carried out to test for DNA sequences that are bound by transcription
factors in vitro (Figure 1-4). Briefly, a tagged transcription factor of choice is purified,
applied to a double stranded DNA microarray, probed with a labelled antibody that
recognizes the protein tag and DNA sequences bound by the protein identified based on
positions on the array that are illuminated (Berger and Bulyk, 2006; Berger et al., 2006). For
these experiments, DNA microarrays have been produced that represent all possible DNA
sequence variants of 10 base pair binding sites (Berger et al., 2006).
18
Figure 1-4: Schematic of a protein binding microarray experiment. Tagged proteins are purified and applied to a microarray with all combinations of 10-mer binding sites. Transcription factor-DNA sequence interactions are identified by treating the microarray with an antibody conjugated to a fluorophore.
19
The universal design of this microarray allows protein binding experiments to be
carried out across different organisms and proof-of-principle of this approach was shown by
identifying consensus binding sites for five transcription factors from yeast (Cbf1 and Rap1),
worm (Ceh-22), mouse (Zif268) and human (Oct-1) (Berger et al., 2006).
Prior to 2008, a catalogue of DNA binding specificities of transcription factors in an
organism was lacking. Even in yeast, sequence motifs for only about half of the ~200
transcription factors were known. To better characterize sequence motifs bound by
transcription factors, one study reported sequence specificities for 112 yeast DNA binding
proteins (Badis et al., 2008). The majority of these binding specificities were derived from
carrying out protein binding microarray experiments on purified DNA binding domains for
transcription factors. Of the 112 sequence specificities identified, 63 motifs were previously
known and match the specificities known for these proteins (Badis et al., 2008). These
results validated the PBM approach and also defined a number of sequence motifs for
transcription factors that were previously unknown. Lending further support to this, many
sequence motifs identified from this study are found upstream of functionally related genes,
suggesting transcription factors co-ordinately control expression of these genes for specific
processes (Badis et al., 2008).
Shortly after this publication, another paper was published using the same PBM
approach to identify direct binding sequences of 89 transcription factors (where 157 proteins
did not yield binding profiles), revealing 50 new DNA binding site motifs in yeast (Zhu et
al., 2009). Both studies identified 2 proteins (called Pbf1 or Ybl054w and Pbf2 or Dot6) that
bind to the PAC and RRPE motifs (Badis et al., 2008; Zhu et al., 2009), which are binding
sites found in promoters of ribosomal RNA genes (Hughes et al., 2000b). Zhu et al. used
20
expression microarrays in a pbf1Δpbf2Δ double mutant to show that ribosomal target genes
of Pbf1 and Pbf2 are no longer repressed under heat shock conditions indicating Pbf1 and
Pbf2 are in fact regulators of these genes (Zhu et al., 2009). Additionally, both Pbf1 and
Pbf2 localized to a subset of PAC-containing ribosomal processing regulatory regions by
ChIP (Zhu et al., 2009). Together the studies by Badis et al. and Zhu et al. significantly
advance our knowledge of cis-binding sequences that are bound by yeast transcription factors
and will guide future studies to examine transcription factor localization at promoters
containing specific binding sites under a series of conditions in vivo.
1.4 Genome-wide nucleosome occupancy
To pack the large genomes of eukaryotic organisms into the nucleus of the cell, DNA
is organized into repeating units called nucleosomes that form chromatin. Histones are the
protein components of nucleosomes that form the histone octamer, which is built from two
histone H2A-H2B dimers and a histone H3-H4 tetramer. 147 base pairs of DNA are
wrapped around the histone octamer to form a nucleosome, which are present in ordered
arrays along the DNA and are separated by linker DNA.
In S. cerevisiae, the PHO5 promoter has been studied as a model to understand the
influence nucleosome occupancy has on transcriptional output. The PHO5 gene encodes an
acid phosphatase and is actively transcribed when intracellular phosphate levels are low and
repressed when phosphate levels are high (Tait-Kamradt et al., 1986). The PHO5 promoter
contains two upstream activating sequences (UAS) that are bound by the transcriptional
activators Pho2 and Pho4 (Barbaric et al., 1996). An important aspect underlying activation
of PHO5 transcription is that nucleosome disassembly must occur to allow activation and
21
this disassembly is mediated by the histone H3/H4 chaperone protein Asf1 (Adkins et al.,
2004; Boeger et al., 2003; Reinke and Horz, 2003). Interestingly, genetic evidence suggests
Asf1-mediated nucleosome disassembly is not required for binding of Pho2 or Pho4 to the
UAS sites in the PHO5 promoter (Adkins et al., 2007). Instead, ChIP analysis of TATA
binding protein (TBP) and RNA polymerase II (Pol II) under activating conditions at the
PHO5 promoter revealed that in the absence of Asf1, TBP and Pol II no longer localize to
the promoter (Adkins et al., 2007). These data indicate that chromatin disassembly at the
PHO5 promoter is required to recruit the general transcriptional machinery to allow activated
transcription. In addition, Asf1 and thus chromatin disassembly is also required for
recruitment of SWI/SNF and SAGA (Adkins et al., 2007), two protein complexes that
modify chromatin and also play a role in activation of PHO5 transcription. Interestingly,
Asf1 appears to be present at many locations throughout the genome (Adkins et al., 2007;
Schwabish and Struhl, 2006), indicating it may be poised on chromatin waiting for specific
transcriptional activators to bind their respective cis-regulatory sites to activate transcription
(Adkins et al., 2007). These PHO5 studies exemplify the role nucleosome assembly and
disassembly play in proper gene regulation, indicating that knowledge of the positions of all
nucleosomes in wild type cells would be informative for better understanding of global
transcriptional control.
Identifying the positions of nucleosomes on DNA involves carrying out a protection
assay where DNA wrapped around the histone octamer is protected from micrococcal
nuclease digestion so that linker DNA separating nucleosomes is preferentially digested
(Figure 1-5) (Yuan et al., 2005). After microccocal nuclease digestion, DNA is purified
from nucleosomes and analyzed to determine specific regions of DNA that were wrapped
22
around the histone octamer. For large-scale assessment of genomic regions occupied by
nucleosomes, high-resolution tiling microarrays are utilized. One study used a microarray
with 50-mer DNA probes tiled every 20 base pairs spanning chromosome III in S. cerevisiae,
allowing positions of nucleosomes to be determined with 20 base pair resolution (Yuan et al.,
2005).
23
Figure 1-5: Schematic representation of a genome-wide nucleosome occupancy experiment. Nucleosomes are cross-linked to DNA with formaldehyde, and nucleosomal DNA is prepared. Micrococcal nuclease treatment preferentially digests linker DNA compared to protected nucleosomal DNA. Cross-linking is reversed and DNA is treated with DNase I to digest DNA to ~50 base pair fragments. To identify genomic regions protected from micrococcal nuclease digestion, DNA is hybridized to a high resolution DNA tiling array and signal is normalized to the hybridization signal from genomic DNA hybridization. Green spots represent genomic regions occupied by nucleosomes while white spots are regions devoid of nucleosomes, as long as genomic DNA is present.
24
The first complete genome map of nucleosome positions was published in 2007
where an even higher density tiling array was used (Lee et al., 2007). This array contains the
S. cerevisiae genome tiled at 4 base pair resolution (David et al., 2006). This study defined
positions of ~71,000 nucleosomes spanning 81% of the genome (Lee et al., 2007) and
revealed that 87% of transcribed genes are occupied by nucleosomes. Intergenic regions
tend to be depleted where only 53% encompass nucleosomes, a trend that is consistent with
ChIP-chip experiments assaying histone levels across the genome (Pokholok et al., 2005).
The general trend of nucleosome-depleted promoters compared to occupied transcribed
regions has been reported by other groups using approaches with different degrees of
resolution, adding to the validity of the data presented by Lee et al. (2007) (Bernstein et al.,
2004; Lee et al., 2004; Yuan et al., 2005). Nucleosome-depleted regions (NDRs) are
generally found 50 base pairs upstream of the transcriptional start site (TSS). Interestingly,
this finding is also supported using a different technique called formaldehyde assisted
isolation of regulatory elements (FAIRE) to positively select regulatory elements (Giresi and
Lieb, 2009; Nagy et al., 2003). Briefly, in FAIRE, protein-DNA interactions are cross-linked
with formaldehyde, sonicated to shear chromatin and treated with phenol-chloroform (Nagy
et al., 2003). DNA devoid of histones separates into the aqueous phase which can be
detected by hybridization of this DNA to a microarray or by sequencing. FAIRE DNA was
found to contain promoter regions upstream of genes, suggesting that these regions are in
fact devoid of histones (Nagy et al., 2003), the most abundant proteins on DNA with the
highest cross-linking efficiency (Brutlag et al., 1969; Polach and Widom, 1995; Solomon and
Varshavsky, 1985).
25
Promoters of actively transcribed genes tend to contain characteristic NDRs
compared to promoters of genes expressed at lower levels which have higher nucleosome
occupancy. Interestingly, open reading frames show an opposite relationship where highly
transcribed genes are occupied with nucleosomes compared to ORFs expressed at low levels
(Lee et al., 2007), possibly because transcriptional initiation and elongation requires ordered
nucleosomes (Lee et al., 2007). Nucleosome depleted regions at promoters correlate with
transcriptional output. For instance, promoters of genes likely not expressed under standard
growth conditions, such as stress-responsive genes, lack nucleosome free regions upstream
of the TSS (Lee et al., 2007). However, promoters of genes that are actively transcribed, like
genes required for protein translation and ribosome production, have characteristic
nucleosome depleted regions upstream of the TSS (Lee et al., 2007). The nucleosome
depleted regions often contain transcription factor binding sequences that in many cases are
80-100 base pairs upstream of the TSS. Examples of transcription factors that have binding
sites at many promoters at the NDRs are Abf1, Reb1 and Mbp1 (Lee et al., 2007).
Many factors contribute to the proper positioning and occupancy of nucleosomes on
DNA including chromatin remodelers, transcription factor binding and the DNA sequence
itself. Because in vivo studies in wild type cells report on the contribution of all these
factors, in vitro studies were carried out to understand the sole contribution of DNA
sequence on nucleosome occupancy. Briefly, genomic DNA was purified from yeast,
incubated with purified chicken erythrocyte histone octamers and then subject to micrococcal
nuclease digestion (Kaplan et al., 2009). The protected nucleosomal DNA was identified by
sequencing to determine the positions on DNA where nucleosomes assembled and relative
occupancy at each base. This experiment defined the DNA sequence preference of
26
nucleosomes and, upon comparison to in vivo maps, was used to determine the contribution
of DNA sequence to nucleosome positions in living cells (Kaplan et al., 2009). A high
correspondence between in vitro and in vivo maps was discovered (correlation of 0.74)
indicating that DNA sequence preferences of nucleosomes play a dominant role in their
occupancy (Kaplan et al., 2009). Higher correlation between in vivo and in vitro maps was
found at non-promoter intergenic regions compared to promoters and open reading frames,
also indicating that DNA sequence is not the sole determinant of nucleosome organization
(Kaplan et al., 2009). To derive a general set of rules governing sequence-determined
nucleosome occupancy, a computational model trained from the in vitro nucleosome data
was developed (Kaplan et al., 2009). The model could predict in vivo nucleosome
occupancy in yeast and in C. elegans (albeit to a lesser degree), indicating the usefulness of
the in vitro yeast data in predicting DNA features that mediate nucleosome organization in
other organisms.
Nucleosome positions were also monitored in living cells in different growth
conditions that are known to change global transcription patterns. When comparing
nucleosome positions from cells grown in glucose (standard growth conditions), galactose or
ethanol, global nucleosome occupancy was found to correlate well amongst the three maps
and these maps also showed a high correlation with the in vitro map (although gene-specific
differences were also observed) (Kaplan et al., 2009). These results indicate that
nucleosomes have an intrinsic preference for DNA sequences and that this preference plays a
central role in establishing nucleosome organization.
27
1.5 The yeast deletion array and the synthetic genetic array (SGA) approach
Assigning functions to genes classically involves cataloguing phenotypic outcomes of
gene perturbations. Typically, forward genetic approaches have been employed where
random mutagenesis of cells to perturb genes has been coupled with follow-up genetic
approaches to characterize mutant phenotypes. Although this approach has proven extremely
useful, random mutagenesis does not provide a saturating survey of all genes and
identification of mutant genes often involves considerable work.
Since genomes of a variety of organisms are fully sequenced, researchers have
knowledge of all predicted open reading frames (ORFs). These gene models have allowed
the creation of genome-wide resources where all genes are deleted or can be systematically
perturbed. RNA interference (RNAi) is typically used to knockdown gene function by
reducing levels of messenger RNA (mRNA). In Caenorhabitis elegans, Drosophila
melanogaster, human and mouse, genome-wide RNAi libraries exist to reduce expression of
each gene (Boutros et al., 2004; Kamath et al., 2003; Moffat et al., 2006). For example, in
human HT29 colon cancer cells, RNAi-mediated knockdown of gene expression was
combined with high-content imaging to identify genes required for proper regulation of
mitosis (Moffat et al., 2006). Measurements of histone H3 phosphorylation on serine 10
(which marks mitotic cells), DNA content by Hoechst staining and actin organization by
phalloidin staining, were combined with RNAi suppression of gene expression to reveal
several known genes and ~100 novel genes associated with mitotic progression (Moffat et
al., 2006).
In C. elegans, RNAi has been used to examine the phenotypic consequences of
perturbation of single genes (Kamath et al., 2003) and to explore genetic interactions. In C.
28
elegans, systematic RNAi screens can be performed very simply by feeding worms bacteria
expressing constructs that produce double-stranded RNA designed to target every worm gene
(Timmons and Fire, 1998). This astounding feature of the C. elegans model means that the
phenotypic consequence of combining an RNAi library with any query mutation can be
assessed. In one study, 37 query mutations of genes involved in cellular signalling pathways
were combined with an RNAi library targeting ~1750 genes involved in signal transduction,
transcriptional regulation and chromatin remodeling producing pairwise combinations of
~65,000 genes (Lehner et al., 2006). This study identified 350 genetic interactions amongst
signalling genes and library genes, many of which are perturbed in human diseases (Lehner
et al., 2006). A more recent genetic interaction study using RNAi identified more genetic
interactions in C. elegans than the Lehner et al. study, even though fewer pairwise
combinations of gene mutations were tested (Byrne et al., 2007). These results suggests that
the spectrum of genetic interactions in the worm may be much larger than initially expected.
In budding yeast, more directed approaches have been used to generate strain
collections where each full length gene is deleted. Here, researchers take advantage of the
high intrinsic rate of recombination to replace each gene with a dominant antibiotic resistant
cassette. The first construction of a complete gene-deletion library was reported in budding
yeast, Saccharomyces cerevisiae, where a kanamycin resistance cassette (KanMX) with
flanking regions homologous to each yeast ORF was targeted to replace each yeast coding
sequence (Figure 1-6) (Giaever et al., 2002). Gene-replacement strategies have been used to
create deletion libraries in other organisms like Schizosaccharomyes pombe (available from
BiONEER), E. coli (Baba et al., 2006) and Cryptococcus sp (Idnurm et al., 2009).
29
Figure 1-6: Gene-deletion strategy for replacing each yeast ORF with a kanamycin resistant cassette (KanMX). The gene knockout cassette is transformed into yeast cells and is precisely integrated in the genome by homologous recombination. Each deletion cassette has up and down (DN) tags which contain unique DNA sequences or barcodes that can be used for strain identification (Giaever et al., 2002; Winzeler et al., 1999).
30
Characterizing the library of yeast deletion strains revealed that approximately 20%
of all 6000 yeast genes are essential for haploid viability, only about half of which were
previously known (Giaever et al., 2002). This remarkable result revealed that extensive
redundancy exists in the genome that buffers the consequence of single gene-deletions. This
observation fuelled development of a high-throughput method for genetic manipulation of
the yeast deletion library termed the synthetic genetic array (SGA) approach (Figure 1-7)
(Tong et al., 2001; Tong et al., 2004). Initially, this approach was used to combine a query
mutation of choice with each yeast deletion strain to create an output array of double mutants
where arrayed colonies could be scored for a growth defect that is more severe than each
single gene deletion (so called synthetic lethal or synthetic sick interactions). Key features of
the SGA approach are: (1) it is automated by use of robotics to replicate arrays of yeast
colonies onto different selection media; (2) since yeast deletion mutants are arrayed, the
position of each yeast mutant is known; (3) yeast of MATα mating type with a mutation in
gene ‘A’ can be mated to the array of MATa yeast deletion mutants to produce diploid yeast
strains; (4) the arrayed diploid strains can be sporulated, and (5) MATa meiotic progeny with
both marked gene deletions can be preferentially selected because they contain a STE2
promoter driving expression of an auxotrophic marker gene that is only expressed in MATa
cells, allowing them to grow on media lacking that particular amino acid (Tong et al., 2001;
Tong et al., 2004).
31
Figure 1-7: The synthetic genetic array approach used for high-throughput double mutant strain construction. A query mutation in a particular ORF (orfAΔ) marked with a nourseothricin resistant cassette (Nat) is mated to the array of viable haploid deletion mutants using a robotic replica pinning procedure. Diploid yeast strains are selected and sporulated and mieotic haploid progeny with the query mutation combined with each yeast deletion mutant is selected. Genetic interactions are identified by observing double mutants that have a growth defect that is more severe than the product of each single mutant. See Tong et al., 2001.
32
The SGA protocol produces an output array of haploid strains where the query
deletion in gene ‘A’ is combined with each deletion mutant which can subsequently be
assayed for double mutant strains that show a synthetic lethal or synthetic sick phenotype.
This approach was applied to probe the yeast genome for genes that were inviable when
combined with 8 query mutations which revealed 291 interactions among 204 genes (Tong et
al., 2001). In a follow-up study, 132 query mutations were screened against the array of
deletion mutants to generate a larger network charting ~4000 synthetic lethal relationships
among ~1000 genes (Tong et al., 2004). This screening approach revealed previously
unappreciated relationships among genes and pathways that span the spectrum of cellular
processes and provides a powerful means to assign gene function.
The SGA-based strain construction strategy provided the tools for another high-
throughput strategy to produce double mutant strains for profiling genetic interactions called
diploid-based synthetic lethality analysis on microarrays or dSLAM (Pan et al., 2004). This
approach involves introduction of a query mutation of choice by transformation into a pooled
collection of the heterozygous MATα/a deletion collection (Giaever et al., 2002) and
conversion of this collection into MATa haploid double mutants that now have the query
mutation combined with each deletion mutant (Pan et al., 2004). As noted earlier, the yeast
deletion collection contains unique strain identifiers or molecular barcodes that flank the
KanMX knockout cassette (Giaever et al., 2002 and Figure 1-6). The barcodes can be PCR
amplified from a pool of mutants and the resulting DNA hybridized to a DNA microarray
with probes complementary to each barcode as a means to identify deletion strains present in
the pool. In dSLAM, double mutants that cause a growth defect or lethality compared to
each single mutant will drop out of the pooled population and those barcodes will be under-
33
represented on the microarray compared to the single mutant, thus identifying genetic
interactions among genes (Pan et al., 2004). This approach was recently used to identify
genetic interactions among genes involved in histone acetylation and deacetylation, revealing
new functional roles for histone acetyltransferases (HATs) and histone deacetylases
(HDACs) (Lin et al., 2008).
Although the SGA approach was initially described to genetically manipulate the
array of viable haploid deletion strains, other genetic arrays can also be screened. For
instance, the functions of essential genes can be probed using SGA by screening arrays of
tetracycline-repressible alleles (Davierwala et al., 2005; Mnaimneh et al., 2004), temperature
sensitive strains and alleles which contain a disruption in the 3’ untranslated region (UTR) of
essential genes to knockdown gene expression (called the decreased abundance of mRNA by
perturbation or DAmP strains) (Schuldiner et al., 2005). Other types of genome-wide arrays
are available that allow overexpression of each yeast ORF, which again often impinges on
cellular growth, particularly when combined with sensitized genetic backgrounds (Sopko et
al., 2006). In addition to screening other arrays in S. cerevisiae, high-throughput strain
construction methodologies have been extended to other organisms including
Schizosaccharomyces pombe (Dixon et al., 2008; Roguev et al., 2007) and E. coli (Butland et
al., 2008; Typas et al., 2008).
1.6 Reporter-based screens
Above, I describe a number of genome-wide approaches largely based on DNA
microarray technology that have been employed to make major discoveries on global gene
regulation. Other major insights have derived from a variety of genetic screening approaches
34
designed to discover regulators of a particular gene of interest. Reporter genes have been
particularly useful in this regard, and I provide an overview of reporter gene screens in yeast
below, since they form the basis of the functional genomic approach that I developed
(Chapter 2).
Reporter genes are constructed by fusing a promoter or any cis-regulatory element of
interest to a gene that provides an easily assayable readout of the activity of that particular
promoter (or cis element). One such reporter gene that has been widely employed is the lacZ
gene from E. coli that encodes the enzyme β-galactosidase. When a promoter-lacZ reporter
gene is introduced into a particular cell, the activity of the promoter can be assayed by
plating cells on media containing X-gal. When β-galactosidase is produced, the substrate X-
gal is cleaved which results in the formation of blue colonies. If the promoter is inactive in
that particular cell, lacZ expression is turned off and colonies appear white.
Reporter technology can be combined with forward genetic screens to isolate mutants
that cause differential expression of the reporter gene, which elucidates candidate regulators
of the promoter being analyzed (Figure 1-8). Randomly mutagenized cells that result in
increased or decreased reporter activity are isolated and the gene mutation characterized to
discover the gene responsible for controlling the promoter driving lacZ expression.
35
Figure 1-8: A forward genetic reporter screen to identify regulators of a promoter of interest. A strain that harbours a promoter-lacZ reporter gene is mutagenized. Mutants are screened for a defect in lacZ production on media containing X-gal. Higher lacZ levels produce dark blue colonies and mapping of the mutation reveals a repressor of the promoter driving lacZ expression. Lower lacZ levels produce white colonies and mapping of the mutation reveals an activator of the promoter driving lacZ expression.
36
An excellent example of early applications of this approach is the effort to learn how
the HO gene in S. cerevisiae is regulated. The HO gene encodes an endonuclease that is
involved in mating type switching. HO transcription occurs only in haploid MATa or MATα
cells but not in a/α diploid cells (Jensen et al., 1983). Furthermore, HO expression is cell
cycle-regulated with transcripts peaking in late G1 and HO expression is restricted to the
mother cell and is not seen in the newly formed daughter cell (Nasmyth, 1983). A forward
genetic screen was employed to identify trans-acting factors that control expression of a HO
promoter-lacZ reporter gene (Breeden and Nasmyth, 1987). Five new SWI genes were
identified that caused a defect in HO promoter-lacZ transcription (Breeden and Nasmyth,
1987), adding to the five already known (Stern et al., 1984). Further dissection of the HO
upstream regulatory sequence (URS) revealed that a short motif repeated throughout the
URS, now called the Swi4-6-dependent cell cycle box (SCB) element [CACGA]4 is
sufficient to confer cell cycle regulation of the HO gene (Breeden and Nasmyth, 1987). To
characterize which SWI mutants specifically act through the [CACGA]4 motif in the HO
promoter, a [CACGA]4-lacZ reporter gene was introduced into the 10 SWI mutants. Only
SWI3, SWI4 and SWI6 deletion strains caused a defect in [CACGA]4-lacZ expression,
suggesting these genes are acting specifically through this cis-regulatory sequence while
other activators like SWI1, 2 and 5 regulate the promoter independently of [CACGA]4
(Breeden and Nasmyth, 1987). Further work revealed that Swi4 and Swi6 directly interact
through their C-terminal regions to form a heterodimeric transcription factor called SBF and
that Swi4 is responsible for directly binding DNA (Andrews and Herskowitz, 1989; Andrews
and Moore, 1992). It was also discovered that chromatin remodelers such as SWI/SNF and
37
SAGA are required for proper regulation of HO transcription and that the protein Ash1
represses HO transcription in daughter cells.
Many of the factors that affect HO were initially discovered using reporter gene
technology but the order of events that lead to HO activation remained unclear. Chromatin
immunoprecipitation studies were carried out to clarify the timing and order of these
transcriptional regulators at the HO promoter. These experiments revealed the following
series of events: (1) Swi5 arrives at the promoter in late anaphase, (2) Swi5 localization to
the HO promoter then recruits SWI/SNF, (3) SWI/SNF recruits the SAGA complex and (4)
recruitment of SWI/SNF and SAGA is required for recruitment of SBF (Cosma et al., 1999).
These events allow HO transcription to occur in the mother cell while the Ash1 protein
represses HO transcription specifically in the daughter cell. These results, largely motivated
by applications of reporter genes, demonstrated complex regulation of a promoter that is
under combinatorial control and shows lineage-specific transcription.
Reporter gene screens have become a standard tool in transcriptional analysis.
Normally, forward genetic screening approaches are employed by randomly mutagenizing
strains harbouring the reporter gene of interest but limitations to this approach exist. First,
random mutagenesis is normally not saturating for the genome nor is it truly random.
Second, a large mutant library must be constructed to ensure high coverage of the genome.
Third, a large amount of follow-up work is required to characterize mutants. Fourth, these
types of reporter screens provide only qualitative measures of gene expression.
To combat many of these problems, we recently described methods for carrying out
systematic reverse genetic screens to identify regulators of a promoter of interest using
functional genomic tools and resources in yeast (Sassi et al., 2009). We describe a system
38
where a wild type yeast strain carrying a reporter gene of interest can be introduced into an
ordered array of ~4500 yeast deletion mutants using the SGA procedure described above.
The use of two types of reporter genes are described, which are based on a colourimetric
assay or auxotrophy. In one case, the activity of a promoter fused to lacZ could be assayed
in each yeast deletion mutant by replica plating the entire array onto medium containing the
substrate X-gal (see above). On this medium, deletion mutants that are required for
repression of the promoter driving lacZ expression will result in higher levels of lacZ
transcription and thus greater β-galactosidase activity leading to very blue colonies. If the
deletion mutant is an activator of the promoter, lacZ transcription will be reduced leading to
a white colony colour.
A second reporter gene is based on HIS3 auxotrophy. Cells that cannot produce
histidine because they lack the HIS3 gene must be grown on media supplemented with
histidine. In this case, a reporter gene harbouring a promoter fused to HIS3 is introduced into
the yeast deletion array using the SGA methodology and deletion of genes that allow growth
on medium lacking histidine are scored as repressors of that promoter (Sassi et al., 2009).
This type of screen is useful only if the promoter driving HIS3 is weak enough so that a
growth defect is seen when cells are grown on medium lacking histidine or a mutation is
made so that the promoter driving HIS3 is inactivated.
The HIS3 reporter gene approach was used to screen for new regulators of the SCB
element, which is bound by the transcription factor SBF (see above). In this case, an SCB-
HIS3 reporter gene was introduced into a strain with a CLN3 deletion (Costanzo et al., 2004).
The absence of the cyclin CLN3 prevents activation of SCB-dependent transcription, thus
cells cannot grow in the absence of histidine because of the failure of the SCB element to
39
drive transcription of the HIS3 gene. The SGA methodology was used to combine the
reporter gene and CLN3 deletion with each yeast deletion mutant. The resulting array of
yeast mutants was screened on media lacking histidine, and it was found that deletion of
WHI5 resulted in growth of colonies because the absence of Whi5 relieved repression of the
SCB element (caused by the CLN3 mutation) and allowed HIS3 transcription (Costanzo et
al., 2004). Follow up experiments revealed that Whi5 association with SBF represses
transcription but upon CDK phosphorylation, Whi5 dissociates from SBF to allow
expression of late G1 genes (Costanzo et al., 2004). This study defined a pathway in yeast
analogous to the Rb-E2F pathway in mammalian cells, which is often targeted in many types
of tumours.
1.7 Global cell cycle transcription
My thesis work has involved the systematic analysis of pathways leading to
activation of cell cycle transcription in yeast. In eukaryotic organisms, the mitotic cell cycle
is an ordered series of events where one cell gives rise to a daughter cell with identical DNA
content. The daughter cell undergoes a gap phase (G1) then upon reaching a critical cell
size, commits to another round of cell division at a point called START in yeast or the
restriction point in mammalian cells. After G1, cells synthesize their DNA (S-phase),
undergo a second gap phase (G2) then ultimately produce a daughter cell after completing
mitosis (M phase). Many input signals underlie cell cycle progression like nutrient
availability, cell size, transcription and protein production and degradation. Proteins called
cyclins are the regulatory components of cyclin dependent kinases (CDKs) that are critical
for the proper control of cell cycle events. Cyclins were so named because their transcript
40
and protein levels peak at particular cell cycle phases and are rapidly degraded when cells
transit outside of that phase (Evans et al., 1983). This oscillation of cyclin levels ensures that
bursts of CDK activity occur at critical points of the cell cycle. For example, in yeast the
CDK Cdc28 associates with G1 cyclins (Cln1, 2 and 3) to promote progression through G1
of the cell cycle. Alternatively, Cdc28 associates with B type cyclins (Clb5 and 6) to
promote replication of DNA while association with Clb1, 2, 3 and 4 is required for proper
progression through G2 and M phases of the cell cycle (reviewed in Mendenhall and Hodge,
1998).
In addition to cyclins, many other gene transcripts fluctuate throughout the cell cycle.
For example, histone genes are among the earliest discovered cell cycle-regulated transcripts
(Hereford et al., 1981). Histone gene expression is tightly regulated so that the genes are
transcribed at high levels during S-phase of the cell cycle when histones are needed to meet
the demands of DNA replication. In the late 1990’s, gene expression microarrays were
utilized to discover the complement of genes whose transcripts are cell cycle-regulated.
Briefly, synchronized yeast cultures were grown so that samples could be taken at fixed time
points to cover each phase of the cell cycle and RNA from cells taken at these time points
was analyzed with gene expression microarrays (Cho et al., 1998; Spellman et al., 1998).
These studies revealed that ~400 to 800 yeast genes are cell cycle-regulated, equating to
~10% of the genome (Cho et al., 1998; Spellman et al., 1998). A more recent study was
carried out to probe cell cycle transcription and the combination of these data with previous
studies revealed that upwards of 1000 yeast genes are cell cycle-regulated (Pramila et al.,
2006). Comparable experiments have been carried out in other eukaryotes including three
studies in S. pombe which individually identified 407 periodically expressed transcripts
41
(Rustici et al., 2004), 747 cell cycle transcripts (Peng et al., 2005) and 750 genes with
significant cell cycle oscillations but suggesting as many as 2000 genes might be cell cycle-
regulated (Oliva et al., 2005). In human cell lines similar experiments revealed >1000 cell
cycle-regulated genes (Cho et al., 2001; Whitfield et al., 2002). In comparing cell cycle-
regulated transcription between S. cerevisiae and S. pombe, one study reported 142
transcripts that were cycle-regulated in both organisms (Peng et al., 2005). Overlapping cell
cycle-regulated genes in both organisms tended to be involved in core cell cycle functions
like DNA replication, chromosome maintenance and mitosis defining a core group of cell
cycle genes that seem to be conserved throughout evolution (Oliva et al., 2005; Peng et al.,
2005; Rustici et al., 2004). Differences in cell cycle-regulated transcripts were often seen
with genes involved in cell metabolism and growth or cell wall biogenesis, likely reflecting
differences in the cell cycle biology of S. pombe and S. cerevisiae (Oliva et al., 2005; Peng et
al., 2005). These studies have collectively defined massive waves of gene expression that
underlie cell cycle progression and imply the existence of a transcriptional regulatory
network that controls proper expression of these genes.
Often times, regulators of cell cycle transcription control genes that perform functions
specific to that phase of the cell cycle. For instance, SBF and MBF are sequence-specific
heterodimeric transcription factors that generally regulate genes transcribed at the G1-S
transition (Koch et al., 1993). ChIP-chip studies carried out to identify all regulatory regions
occupied by these transcription factors showed that SBF generally controls genes involved in
budding and membrane and cell wall biosynthesis while MBF for the most part controls
genes that function in DNA replication and repair (Iyer et al., 2001). Additionally, a number
of target genes defined by ChIP-chip are cell cycle-regulated and these were enriched for
42
genes expressed during G1 and S-phase (Iyer et al., 2001). Some target genes showed peak
expression outside of G1 and S-phase. For example, the G2/M cyclins CLB1 and CLB2 were
bound by Swi4, suggesting SBF plays a more complex role than previously thought (Iyer et
al., 2001).
Waves of gene expression at one cell cycle phase are important for proper timing and
expression of genes at the adjacent phase. Also, key regulators of the cell cycle are usually
themselves cell cycle regulated and ChIP-chip studies on 9 of these transcriptional activators
(Mbp1, Swi4, Swi6, Mcm1, Fkh1, Fkh2, Ndd1, Swi5 and Ace2) described a cyclic model
whereby transcriptional regulators drive expression of each other (Simon et al., 2001).
Additionally, multiple activators function at each cell cycle phase, indicating that extensive
buffering exists so that cell cycle progression is not severely impaired by single mutations in
transcription factors. This study set the stage for future work to decipher the complete group
of regulators that control cell cycle transcription. ChIP-chip experiments and analysis of
transcription factor binding sites of co-expressed genes from microarray data, identified the
transcription factor Hcm1 as an important regulator of S-phase specific transcription, filling a
key gap in our knowledge of cell cycle transcriptional regulation in yeast (Horak et al., 2002;
Pramila et al., 2006).
These and other studies led to the general model that cell cycle transcriptional control
is a closed circuit where activators at one phase control expression of transcription factors at
the next phase. In late G1, SBF (Swi4-Swi6) is responsible for activation of Hcm1 which is
active at late S-phase (Horak et al., 2002; Iyer et al., 2001; Pramila et al., 2006). Hcm1, SBF
and MBF account for activation of Ndd1, Fkh1 and Fkh2, transcription factors that together
with Mcm1 regulate G2/M transcription and in turn regulate Swi5 and Ace2 (Horak et al.,
43
2002; Iyer et al., 2001; Pramila et al., 2006; Simon et al., 2001). Swi5-Ace2-Mcm1 controls
M/G1 transcription (Simon et al., 2001). Mcm1 is then responsible for activation of Swi4,
closing the transcriptional circuit and allowing another burst of G1 transcription during the
next cell cycle (Simon et al., 2001). These results set the framework for understanding cell
cycle transcriptional control (key regulators are summarized in Table 1-1), but does not
account for the cell cycle periodicity of hundreds of transcripts indicating many more
discoveries are required to fully understand cell cycle transcription.
44
Table 1-1: Key regulators of cell cycle transcription
Cell Cycle Regulator Activator or Repressor Cell Cycle Phase Regulated SBF (Swi4-Swi6) Activator G1 MBF (Mbp1-Swi6) Activator G1 Hir1 Repressor S Hir2 Repressor S Hir3 Repressor S Hpc2 Repressor S Hcm1 Activator S Fkh1 Activator G2/M Fkh2 Activator G2/M Ndd1 Activator G2/M Mcm1 Activator G2/M, M/G1 Swi5 Activator M/G1 Ace2 Activator M/G1 Yox1 Repressor M/G1 Yhp1 Repressor M/G1
45
A major void in understanding cell cycle transcription is a clear understanding of the
factors and mechanisms that underlie histone transcription that occurs in S-phase. Most of
histone gene regulation has been characterized in terms of repression by the Hir proteins
(Osley and Lycan, 1987). Histone gene expression is a main focus of this Thesis and is
discussed extensively in Chapter 3.
1.8 Summary and overall significance
Above I describe a number of genome-wide approaches that have been broadly
applied to study transcriptional control. Many of these approaches utilize DNA chip
technology, which has revolutionized the field of gene regulation. Parallel analysis of all
transcripts being expressed in a particular cell, locations of regulatory proteins on DNA,
genome-wide positions of nucleosomes, computational prediction and experimental
identification of transcription factor binding sites have led to enormous discoveries on how
the genome is transcribed.
These approaches provide information on how all promoters are being regulated in
the genome and have been combined to identify target genes of particular proteins.
However, some of these approaches require prior knowledge of protein function. For
example in ChIP-chip studies, knowledge of protein function in transcriptional control is
required in order to choose proteins to tag for analysis. By carrying out reporter screens and
using a single promoter as bait, the genome can be analyzed to discover proteins that affect
transcription of that particular promoter. Although extremely useful, advanced techniques
for reporter screening are lacking. Limitations to traditional forward genetic screening
approaches are discussed above. We have recently described a functional genomics
46
methodology by combining lacZ and HIS3 reporter genes with array-based approaches in
yeast to counter some of these problems (Sassi et al., 2009). However, these types of screens
are generally not quantifiable making them difficult to employ in large-scale studies.
In this Thesis, I describe the development and complete methodology of a large-scale
reporter screen which combines available functional genomic tools and resources in yeast
that is unbiased, quantifiable and internally controlled (Kainth et al., 2009 and Chapter 2). I
applied this screening procedure to discover novel regulators of histone gene expression, an
important group of genes that are tightly cell cycle-regulated with peak transcription
occurring during S-phase. I show the utility of this screening approach by identifying a
number of well established regulators of histone gene expression including the Asf1, Hir1, 2
and 3 histone chaperone proteins and the histone periodic control 2 (Hpc2) protein.
Additionally, my screen revealed a novel role for Rtt106, another histone chaperone, in
repression of histone genes as well as a role for the HAT, Rtt109, and the bromodomain
containing protein, Yta7, in activation of histone gene expression. A series of follow-up
experiments are described utilizing ChIP, transcript profiling, genome-wide nucleosome
occupancy and protein-protein interaction studies to show that Rtt106 is a member of the
Hir-Asf1 complex, which is responsible for repression and cell cycle control of histone
genes. This repression is countered in part by Rtt109, likely by acetylating histone H3 lysine
56 (H3 K56Ac). We discovered that this Rtt106 is correctly localized at the promoter
because Yta7 is also present, which restricts Rtt106 from spreading into the coding region of
histone genes. Thus, the work described here represents biological discovery as well as
technology development, which should form the basis of future studies employing reporter
gene screens.
47
Chapter 2
Comprehensive genetic analysis of transcription factor pathways using a dual reporter gene system in budding yeast
The work described in this chapter is published as: Pinay Kainth, Holly Elizabeth Sassi, Lourdes Peña-Castillo, Gordon Chua, Timothy R. Hughes and Brenda Andrews. Comprehensive genetic analysis of transcription factor pathways using a dual reporter gene system in budding yeast. Methods. 48, 2009, 258-264. Permission to reprint this work was obtained from Elsevier. Author contributions:
PK developed the method, analyzed the data, carried out the screening, wrote the manuscript and produced Figures 2-2 through 2-5.
HES helped with development of the method, screening, editing the manuscript and produced Figure 2-1.
LP carried out the data normalization and statistical analysis for data shown in Figures 2-3, 2-4 and 2-5.
GC assisted with developing the method.
TRH assisted with development of the method and editing the manuscript.
BA directed the project and writing of the manuscript.
48
Abstract
The development and application of genomic reagents and techniques has fuelled progress in
our understanding of regulatory networks that control gene expression in eukaryotic cells.
However, a full description of the network of regulator-gene interactions that determine
global gene expression programs remains elusive and will require systematic genetic as well
as biochemical assays. Here, I describe a functional genomics approach that combines
reporter technology, genome-wide array-based reagents and high-throughput imaging to
discover new regulators controlling gene expression patterns in Saccharomyces cerevisiae.
Our strategy utilizes the SGA method to systematically introduce promoter-GFP (Green
Fluorescent Protein) reporter constructs along with a control promoter-RFP (Red Fluorescent
Protein) gene into the array of ~4500 viable yeast deletion mutants. Fluorescence intensities
from each reporter are assayed from individual colonies arrayed on solid agar plates using a
scanning fluorimager and the ratio of GFP to RFP intensity reveals deletion mutants that
cause differential GFP expression. I describe analysis of data from a CLN2 promoter-GFP
screen and show that our method identifies deletion of SWI4, the known activator of CLN2
expression, as causing the greatest defect in CLN2 promoter-GFP fluorescence. The method
is extensible to any transcription factor or signalling pathway for which an appropriate
reporter gene can be devised.
49
2.1 Introduction
With its sophisticated yet straightforward genetic tools, the budding yeast
Saccharomyces cerevisiae has long been a premier model organism for exploring eukaryotic
biology. The utility of yeast for studying conserved processes and biological pathways has
been aided in the past decade or so by access to remarkable genetic resources including a
complete gene deletion collection (Giaever et al., 2002; Winzeler et al., 1999) and
comprehensive arrays of yeast strains carrying tagged alleles of all genes (Huh et al., 2003;
Krogan et al., 2006; Sopko et al., 2006). These resources have allowed yeast to serve as a
test-bed for the development of many functional genomic techniques that are now broadly
applied. Recent advances in our global understanding of transcription factors and their target
genes have relied heavily on gene expression microarrays, genome-wide chromatin-
immunoprecipitation (ChIP-chip) experiments, and computational prediction of transcription
factor binding sites based on these data.
As discussed in Chapter 1, approaches such as ChIP-chip demand prior knowledge of
gene function to choose proteins to tag with the purpose of understanding their roles in
transcriptional regulation at all promoters. Complementary genetic approaches have also
been productively used for many years to discover genes that regulate a promoter of interest.
In particular, the development and use of sensitive reporter genes has enabled many
landmark discoveries about transcriptional regulation over the past two decades. Typically
these types of screens involve fusing a specific promoter to an appropriate reporter gene (pr-
reporter) and assessing the effect of perturbing cis-elements or trans-acting factors on
reporter gene expression. Reporter genes based on the CYC1 promoter and the bacterial lacZ
gene enabled the early definition of the UAS as a key promoter element in yeast (Guarente et
50
al., 1982). Here, promoter activity was assessed by measuring levels of lacZ-encoded β-
galactosidase using simple colourimetric assays. Other types of reporters are based on
nutritional requirements and are detected using a growth readout. For example, a histidine
auxotroph containing a promoter of interest fused to the histidine biosynthetic gene HIS3 can
be screened to identify mutants defective in driving HIS3 expression, which will result in a
growth defect.
As noted above, reporter genes have been extensively used in genetic screens to
discover trans-acting factors that influence expression from a specific promoter leading to
new paradigms of transcriptional regulation. For example, early screens for mutants that
affected expression of an HOpr-lacZ reporter gene led to 1) the discovery of the founding
member of the Swi/Snf family of chromatin remodelers; 2) elucidation of a pathway of
lineage-specific gene expression involving localized mRNA and 3) the description of
transcription factors that induce gene expression during G1 phase, a conserved and important
feature of eukaryotic cell cycles (Andrews and Herskowitz, 1989; Breeden and Nasmyth,
1987; Jansen et al., 1996; Sternberg et al., 1987). These examples illustrate the capacity of
reporter gene screens to uncover both proteins that directly regulate gene expression by
binding to promoters or their regulatory regions as well as upstream factors that influence
their activity.
Although classical genetic screens employing reporter genes have been extremely
useful, they are generally not saturating. Recently, we described an approach that makes use
of the yeast deletion array to analyze expression of lacZ and HIS3 promoter reporter genes
that provides a saturating survey of the effect of loss-of-function mutations in yeast genes on
a promoter of interest (Sassi et al., 2009), the latter of which was used to identify regulators
51
of G1 transcription (Costanzo et al., 2004). However, the readout of these reporters is
difficult to reliably and rapidly quantify. I therefore sought to devise a completely
quantitative and unbiased approach to study transcription factor pathways.
The recent engineering of fluorescent proteins in a variety of different colours across
non-overlapping spectral classes has made multi-colour cell biological experiments feasible
(Shaner et al., 2007; Shaner et al., 2005). Additionally, fluorescent proteins are easy to
detect with the correct optics and fluorescent signals can be rapidly quantified, making them
useful markers of a variety of cellular events.
In this Chapter, I describe a system that combines dual colour promoter-reporter
genes with the SGA methodology (Tong et al., 2001; Tong et al., 2004) to assess the effect of
deleting each yeast gene on a promoter of interest. The system includes a query strain that
harbours any promoter of interest fused to GFP on a plasmid as well as an integrated control
promoter fused to the dsRed variant tdTomato (Shaner et al., 2004) (which we refer to as
RFP). The query strain is crossed to the collection of ~4500 viable haploid yeast deletion
mutants using a simple replica-pinning procedure, resulting in an output array in which each
deletion mutant contains both reporters. By constructing such an output colony array, we can
easily assess the effect of each yeast deletion mutant on reporter gene activity by scanning
both GFP and RFP fluorescence intensities directly from colonies arrayed on agar plates.
After quantifying these data, the GFP intensity captured from each colony can be
standardized to the RFP signal from that same colony to identify deletion mutants causing
differential GFP expression. We expect deletion of a putative activator to result in a
decreased GFP:RFP ratio while deletion of a putative repressor will result in an increased
52
GFP:RFP ratio. This methodology allows a survey of the genome to identify both direct
regulators and upstream signals and pathways that impinge upon a promoter of interest.
2.2 Description of methods
2.2.1 Reporter system
Promoters are PCR amplified from yeast genomic DNA and cloned into the plasmid
BA1926 upstream of a cassette consisting of a GFP reporter followed by the terminator
sequence of the ADH1 gene (Wach et al., 1997) (Table 2-1). BA1926 is a CEN-based and
LEU2-marked plasmid that was derived from the plasmid pRS315 (Sikorski and Hieter,
1989).
In my work, I define a promoter as the intergenic region upstream of a given open
reading frame (ORF) up to 1000 base pairs. The resulting reporter constructs are
transformed into an appropriate SGA query strain harbouring an integrated copy of a control
promoter fused to RFP, marked with the hygromycin resistance gene (HphMX) (Goldstein
and McCusker, 1999) (see below and Table 2-1). Users should be careful to select a control
promoter that is regulated independently of the promoter of interest. Because of my interest
in cell cycle-regulated transcription, I constructed control reporter constructs containing the
promoter of the actin gene, ACT1, or the promoter of the RPL39 gene which encodes a
ribosomal protein. Both of these promoters are constitutively expressed and independent of
cell cycle regulation (Spellman et al., 1998).
53
Table 2-1: Query strains and plasmids used in the two-colour promoter-reporter screening system.
Strain Genotype Source Y7092* MATα can1Δ::STE2pr-his5 lyp1Δ his3Δ1 leu2Δ0 ura3Δ0 met15Δ0 Tong et al.,
2007 BY4256 MATα HO::RPL39pr-tdTomato::hphMX integrated in Y7092 Andrews lab BY4347 MATα HO::ACT1pr-tdTomato::hphMX integrated in Y7092 Andrews lab
Plasmids Description Source pRS315 CEN-based LEU2-marked low copy plasmid Sikorski and
Hieter, 1989 BA1926 GFP followed by ADH1 terminator cloned into pRS315 Andrews lab BA1927 CLN2 promoter cloned upstream of GFP in BA1926 Andrews lab *pr-tdTomato::hphMX reporters were integrated in the SGA compatible query strain Y7092. If other control reporter genes are desired, they should be integrated into the strain Y7092. his5 refers to the gene from Schizosaccharomyces pombe (Tong, 2007).
54
2.2.2 Generating an output array of GFP and RFP reporter genes in yeast deletion mutants
We are using the SGA technique to simultaneously introduce GFP and RFP reporter
genes into an array of ~4500 yeast deletion mutants. The yeast deletion collection (Giaever
et al., 2002) is comprised of strains in which each of the known or predicted ORFs is deleted
and replaced with the KanMX antibiotic-resistance marker (available from OpenBioSystems:
http://www.openbiosystems.com) or EuroScarf: http://web.uni-
frankfurt.de/fb15/mikro/euroscarf).
In Figure 2-1, I summarize the strategy for producing an output array of both reporter
genes in each deletion mutant. The SGA technique automates yeast genetics and has been
described elsewhere, including detailed descriptions of drug selections (Tong, 2005; Tong,
2007). My approach follows the basic SGA protocol, and pertinent features for my method
include: 1) omission of leucine from the media to select growth of yeast carrying the GFP
reporter plasmid and hygromycin B and G418 addition to allow for selective growth of yeast
carrying the RFP reporter gene and gene deletion respectively; 2) omission of histidine to
allow selective growth of MATa meiotic progeny, since only these cells express the STE2pr-
his5 reporter. Other drug additions and amino acid omissions are used to specifically enrich
for haploid yeast after sporulation which is described in detail in (Tong, 2007).
55
Figure 2-1: Overview of the dual reporter SGA methodology. A lawn of the MATα query strain BY4256 or BY4347 containing the LEU2-marked pr-GFP plasmid and control pr-RFP gene is crossed to an array of MATa haploid gene-deletion mutants on YEPD. can1Δ and lyp1Δ markers and canavanine and thialysine selections are described in Tong et al., 2007. The result of the SGA-based cross is an array of yeast deletion mutants, each containing the pr-GFP reporter and RPL39pr-RFP control gene. Fluorescence intensities are measured directly from the colonies arrayed on the final agar plates using a scanning fluorimager and the resulting scans are quantified using GenePix Pro version 3.0 software. Yellow circles represent colonies in which GFP and RFP fluorescence intensities are relatively equal, as is expected for deletion mutants that do not affect GFP expression. Green circles represent yeast strains containing deletions in candidate repressors. Red circles represent yeast strains containing deletions in candidate activators. Digital images of the final selection plates are taken after scanning and colony sizes are scored to exclude strains that are sick, dead, or mis-pinned from the analysis. XXX: wild-type allele, xxxΔ: gene-deletion allele, black oval: LEU2-marked plasmid, blue box: promoter, green box: GFP, SGA methodology adapted from (Tong, 2007).
56
The following serves as a guide to generate haploid yeast strains harbouring LEU2-
marked GFP reporter plasmids and an integrated HphMX marked RFP reporter combined
with a KanMX gene deletion. All manipulations of yeast arrays can be performed using
hand pinning tools or robotics (specialized robots for manipulating yeast arrays are available
from BioRad, Singer Instruments or S&P Robotics). The general protocol summarized
below can be completed in less than three weeks and effectively substitutes for many
thousands of individual yeast transformations.
Day 1: Create a lawn of the strain BY4256 or BY4347 (Table 2-1) harbouring a pr-GFP
reporter gene of interest on a CEN-based plasmid on SC -Leu solid media (prepared in Nunc
brand single well OmniTrays). Grow for 2 days at 30 °C.
Day 3: Mate the lawn to 14 plates of arrayed yeast deletion mutants, in 768-colonies per
plate density on YEPD solid media. Incubate for 1 day at 30 °C.
Day 4: Replicate onto SC -Leu +200 µg/ml G418 (G418 Sulfate/Geneticin, Gibco,
Invitrogen Corp.) diploid selection plates. Incubate at 30 °C for 2 days.
Day 6: Repeat selection on SC -Leu +200 µg/ml G418 plates to further enrich for
heterozygous mutant strains carrying the pr-GFP reporter. Incubate at 30 °C for 1 day.
Day 7: Replicate the array of diploids onto enriched sporulation media. Incubate for 5 days
at 22 °C.
Day 12: Select MATa haploids on SC -Leu -Arg -His -Lys +50 µg/ml canavanine +50 µg/ml
thialysine +200 µg/ml G418 plates. Incubate at 30 °C for 3 days. Canavanine (L-
Canavanine Sulfate Salt) and thialysine [(S-(2-Aminoethyl)-L-cysteine hydrochloride/L-4-
Thialysine hydrochloride)] are available from Sigma Aldrich Co.
57
Day 15: Select MATa haploids on SC -Leu -Arg -His -Lys +50 µg/ml canavanine +50 µg/ml
thialysine +200 µg/ml G418 +300 µg/ml hygromycin B (MP Biomedicals) plates. Grow for
2 days at 30 °C. Addition of hygromycin B allows selection of the control RFP reporter gene
in haploid strains carrying the GFP reporter and deletion mutation. Omitting hygromycin B
from the media on Day 12 conserves use of this drug.
Day 17: Select MATa haploids on final selection SC –Leu –Arg -His -Lys +50 µg/ml
canavanine +50 µg/ml thialysine +200 µg/ml G418 +300 µg/ml hygromycin B solid media.
Grow for 2 days at 30 °C. This step facilitates further enrichment of the final haploid strains.
The Day 17 plates are poured to a volume of 60 ml per OmniTray, generating plates of
optimal thickness for colony fluorescence imaging (see section 2.2.4).
Day 19: Digital imaging of plates (section 2.2.3) and colony fluorescence detection (section
2.2.4).
2.2.3 Digital imaging of yeast plates
The 14 plates representing the output array are individually photographed with a
Canon powershot G2 4.0 megapixel digital camera operated by Remote Capture software
version 2.7.5.27. These images are uploaded into Qt ColonyImager software version 1.01
(Boone and Andrews labs, unpublished software) to derive colony diameter values for each
array position measured in number of pixels. Automated systems for imaging yeast and
bacterial arrays are also available [http://www.sprobotics.com/spImager_Desc.html].
58
2.2.4 Assaying GFP and RFP fluorescence intensities from colonies arrayed on agar plates
GFP and RFP fluorescence intensities are assayed directly from colonies arrayed on 14
agar plates using a fluorescence scanner (Typhoon Trio Variable Mode Imager, GE
Healthcare). The surface of the scanner is large enough to detect fluorescence from 7 plates
in a single run meaning two output scans are required to assay fluorescence signals from the
complete array of 14 plates. To collect colony fluorescence data, lids are removed from the
OmniTrays and the plates are inverted so that colonies are facing the glass surface of the
scanner. Plates are affixed to the scanner surface using clear tape to prevent movement
during the scan. To excite and measure fluorescence intensities from the GFP and RFP
fluorophores, Typhoon Scanner Control version 5.0 software is used with the following
settings:
1. Acquisition mode - fluorescence
2. Laser 1: 488 nm, emission filter: 520 BP40
Laser 2: 532 nm, emission filter: 580 BP30
3. Adjust photomultiplier tube (PMT) voltage so that GFP and RFP intensities measured
from colonies are below saturation. To detect RFP intensity from strains BY4256 and
BY4347, a PMT value of 615 volts is optimal. Users should generate a test plate of
colonies to determine the optimal PMT setting for GFP detection (Laser 1) which will
depend on the strength of the promoter driving GFP expression.
4. Pixel size: 100 microns
5. Focal plane: +3 mm
6. Start scan and save file
59
GFP intensities are acquired first followed by RFP intensities. The duration of each scan is
~25 minutes. Thus, genome-wide scanning of 14 output array plates takes ~50 minutes. We
have taken care to pour final selection plates to a volume of 60 ml using a Wheaton
Unispense media pump so that colonies arrayed on these plates are 3 mm above the surface
of the scanner. Therefore, it is important that a focal plane setting of +3 mm is selected so
that the laser is focused directly on the colonies to allow maximal excitation of fluorescent
proteins in each colony.
2.2.5 Quantifying fluorescence intensities from output scans
The scanned output file (.gel) can be uploaded directly into GenePix Pro version 3.0
software for automated feature detection and quantification of both GFP and RFP intensities.
The median colony fluorescence intensity for each fluorophore corrected for local
background fluorescence is obtained from GenePix and is used for further data normalization
and analysis.
Additional notes:
The yeast deletion array we use is a collection of 14 plates where each deletion
mutant is represented twice. Each plate has 384 deletion mutants arrayed in duplicate in two
separate 16 x 24 grids producing an array with 768 colonies per plate (Tong, 2005). Note
that each plate is bordered by his3Δ deletion strains which benefit from a growth advantage
due to availability of additional nutrients compared to other colonies (Tong, 2007). These
border colonies are removed from further analysis to limit gene expression measurements
60
derived from colonies with favourable growth advantages as a result of their arrayed
position.
2.3 Analysis of data from a genome-wide promoter-reporter screen
I carried out a screen to identify novel regulators of the CLN2 promoter, which drives G1
specific expression of the CLN2 gene, a key cyclin involved in promoting the G1 to S phase
transition (Hadwiger, 1989). CLN2 transcription is activated in late G1 of the cell cycle by
the transcription factor Swi4 (Nasmyth and Dirick, 1991; Ogas et al., 1991). Thus, I expect
deletion of SWI4 to result in decreased CLN2pr-GFP expression compared to its effect on the
control RPL39pr-RFP gene. By visual inspection of an output fluorescence scan, shown in
Figure 2-2, I find that SWI4 deletion does in fact cause a defect in CLN2 transcription.
61
Figure 2-2: Representative fluorescence scan of a single output array plate. The output array consists of yeast deletion mutants carrying the CLN2pr-GFP construct and an integrated RPL39pr-RFP control gene arrayed in duplicate in a 768-colony per plate format. The left panel depicts the GFP fluorescence intensities, the centre panel depicts the RFP fluorescence intensities and the image overlay of both scans is shown in the right panel. Deletion of SWI4, the known hallmark transcriptional activator of CLN2 expression (Nasmyth and Dirick, 1991; Ogas et al., 1991), results in decreased GFP intensity by visual inspection.
62
To identify regulators in an unbiased manner, I use a quantitative approach to generate a
high-confidence list of potential regulators controlling expression from a particular promoter.
This approach is described below in the context of CLN2 transcriptional regulation.
2.3.1 Using colony size data to filter dead/sick colonies from further analysis.
Colony size varies depending on the growth rate of individual mutant strains, which can
confound analysis of colony fluorescence. The data set obtained from each promoter screen
is filtered for dead/sick colonies using colony size values obtained from digital images of the
output array (section 2.2.3). Using a data set of 27 genome-wide pr-GFP reporter screens
(Andrews lab, unpublished data), we determined a threshold of minimal colony diameter
based on the observed distribution of colony size values. Colonies with diameters that
measure less than 345 pixel units (P-value < 0.01, normal distribution) are removed from the
data set. An example of this analysis and colony size distributions for the CLN2pr-GFP
screen is displayed in Figure 2-3.
63
Figure 2-3: Colony size distribution of yeast deletion mutants. Arrays of colonies were imaged after final selection of haploid mutant strains harbouring the CLN2pr-GFP reporter plasmid and the RPL39pr-RFP control gene. Qt ColonyImager version 1.01 software was used to measure colony diameter in pixel units. The x-axis represents colony size measurements from four individual colony replicates of each deletion mutant binned in groups of 10 pixel units. The y-axis represents the frequency with which colony sizes fall into each bin. Based on the observed normal distribution of 27 pr-GFP screens (Andrews lab, unpublished data), we set a minimal colony size threshold of 345 pixel units. Each colony that is below this threshold should be removed from further analysis of fluorescence intensities.
64
2.3.2 Normalization of GFP and RFP intensities
For each screen, the GFP and RFP intensities from both replicate mutant strains are
averaged and GFP:RFP log2 ratios of intensity measurements are obtained. The
normalization algorithm LOESS (Cleveland, 1979) is applied to the log2 ratios. LOESS
seeks to eliminate intensity-dependent artefacts in the data. In addition, since LOESS-
normalized log2 ratios are centered at zero, we can directly compare ratios from different
screens and jointly visualize data collected from multiple screens by clustering and other
computational approaches. We find that the LOESS-normalized log2 ratios of GFP to RFP
are reproducible in replicate screens on different plates and are unaffected by spatial artefacts
and small variations in colony size above the pre-determined threshold described above (data
not shown).
2.3.3 Correlation between replicate screens and display of genome-wide data
I screened the CLN2pr-GFP reporter in duplicate to assess the reproducibility of our
results and found that GFP:RFP log2 ratios of replicate experiments correlated with R=0.90
(Figure 2-4).
65
Figure 2-4: Scatter plot of replicate CLN2pr-GFP screens. We carried out each screen in duplicate to assay the reproducibility of our system and found that log2 GFP:RFP measurements from individual screens correlate with Pearson R=0.90. Correlation was determined by comparing the average log2 GFP:RFP ratio from two replicate colonies from individual screens.
66
Quantified, normalized log2 GFP:RFP fluorescence intensity ratios are sorted in
ascending order to identify deletion mutants with decreased GFP:RFP (putative activators)
and those with increased GFP:RFP (putative repressors) ratios. Figure 2-5 shows the data
obtained from the genome-wide CLN2pr-GFP screen. Points on the curve represent
normalized fluorescent protein intensities averaged from four replicate deletion mutant
colonies. As expected, most deletion mutants do not affect transcription and are represented
on the flat portion of the curve in Figure 2-5A. The sigmoidal nature of the curve suggests
putative activators will be discovered in the tail of mutants with log2 GFP:RFP ratios below
zero and putative repressors identified in the tail of mutants with log2 GFP:RFP ratios above
zero. My survey of the genome reveals that deletion of SWI4, which is the known
transcriptional activator of CLN2 gene expression (Nasmyth and Dirick, 1991; Ogas et al.,
1991), results in the greatest defect in CLN2pr-GFP expression when normalized to the
RPL39pr-RFP measurement (Figure 2-5A). Note that quantitative measurements from a
single reporter screen using only GFP would fail to detect Swi4 as the major activator of
CLN2 gene expression (Figure 2-5B). Instead, Tps2, which is a phosphatase involved in the
synthesis of the carbohydrate trehalose (De Virgilio et al., 1993), would have erroneously
appeared as the top transcriptional activator. This is likely due to the reduced colony size of
the tps2∆ strain on the output array from which the GFP measurement was taken.
In Figure 2-5A, a 2–fold increase and decrease in reporter expression is marked to orient
the reader. To define thresholds for establishing regulators, the data represented here can be
transformed to Z-scores, and P-values can be assigned based on a normal distribution. Users
can set P-value cut-offs for individual screens depending on the stringency desired.
67
Figure 2-5: Screening deletion mutants to identify regulators of the CLN2 promoter. (A) Distribution of log2
GFP:RFP ratios from genome-wide analysis of the CLN2 promoter. The y-axis represents log2 GFP:RFP ratios measured from each deletion mutant displayed on the x-axis. When CLN2pr-GFP fluorescence intensities are standardized to the control RPL39pr-RFP intensities, we find deletion of the known activator SWI4 results in the greatest defect in CLN2 reporter transcription. (B) Quantitative measurements of a single reporter screen. The y-axis represents GFP intensities measured from each deletion mutant displayed on the x-axis. If only CLN2pr-GFP intensities are considered, our quantitative analysis fails to detect Swi4 as the major transcription factor that activates CLN2 gene expression. Instead, we falsely detect Tps2 as the top activator.
68
Additional notes
The majority of our pr-GFP screens were carried out using a CEN-based low copy
plasmid. In cases where GFP expression from a given promoter is weak, the reporter
construct can be moved to a high copy plasmid to sufficiently amplify the GFP signal for
robust detection. Cell-to-cell variation in plasmid copy number should be considered when
designing single-cell imaging experiments. However, our system involves detection of
fluorescence intensities from whole colonies comprised of many cells, making plasmid copy
number of little concern. If desired, it is possible to integrate the GFP reporter gene into the
genome.
It should also be noted that since our reporter constructs contain the entire intergenic
region (up to 1000 base pairs upstream of each ORF), the 5’ UTR corresponding to each
gene being studied will be included in the transcript. Thus, mutants required for post-
transcriptional regulation through interaction with the 5’ UTR could potentially be
indentified in our screen. For example, the 5’ UTR of the Yap2 transcript has been
implicated in decay of YAP2 mRNA (Vilela et al., 1999).
Since our approach relies on the SGA methodology to construct output arrays of
reporter genes in deletion mutant backgrounds, users should be aware that KanMX gene
deletions linked to markers used in the MATα query strain cannot be screened. For example,
screening KanMX deletion strains linked to the HO gene requires integration of the control
promoter-RFP::hphMX gene at a locus other than HO in the MATα query strain. I also note
that because this approach relies on reporter proteins driven by specific promoters, it does
not provide a direct measure of transcript levels of the endogenous gene controlled by the
69
promoter studied. An independent verification assaying endogenous transcript levels in
identified deletion strains should be carried out.
2.4 Concluding remarks
I developed a two-colour promoter-reporter system that makes use of high-throughput
genetics and the array of viable haploid deletion mutants to understand transcription factor
pathways in yeast. This method could also be used to screen any appropriately marked yeast
array, so that additional types of genetic perturbations as well as the roles of essential genes
can be assessed. Relevant strain collections include arrays where each yeast ORF is
overexpressed (Sopko et al., 2006), tetracycline-repressible alleles (Mnaimneh et al., 2004)
and the decreased abundance of messenger RNA by perturbation (DAmP) alleles (Schuldiner
et al., 2005). Additionally, the dual-reporter system can be used to screen promoters in
various experimental conditions by pinning the output reporter gene arrays on the appropriate
medium or growing them under specific environmental conditions. Such an approach should
prove useful in revealing how cells integrate external environmental signals into their gene
expression program. Finally, this two-colour promoter-reporter system can be applied to
virtually any pathway for which a fluorescent reporter can be designed, representing a
powerful and general approach for analysis of biological pathways.
70
Chapter 3
A two-color cell array screen reveals interdependent roles for histone chaperones and a
chromatin boundary regulator in histone gene repression
The work described in this Chapter is published as:
Jeffrey Fillingham*, Pinay Kainth*, Jean-Philippe Lambert, Harm van Bakel, Kyle Tsui, Lourdes Peña-Castillo, Corey Nislow, Daniel Figeys, Timothy R. Hughes, Jack Greenblatt, and Brenda Andrews. Two-colour cell array screen reveals interdependent roles for histone chaperones and a chromatin boundary regulator in histone gene repression. Molecular Cell. 35, 2009, 340-351.
* These authors contributed equally to this work. Permission to reprint this work was obtained from Elsevier. Author contributions: PK developed and carried out the HTA1 promoter-GFP screen, analyzed the reporter data, carried out all of the qPCR experiments, initiated collaborations with JF, JG, KT, CN and HvB, and assisted JF with writing the manuscript. JF carried out all of the ChIP experiments in this study, initiated the collaboration with JL and DF and wrote the initial draft of the manuscript. JL did the co-purification experiment in Figure 3-4D under supervision from DF. KT and CN carried out the nucleosome positioning experiments in Figures 3-5C and D. HvB normalized the nucleosome positioning data and assisted PK with analysis. LP normalized the data from the HTA1 promoter-GFP screen. TRH and JG assisted with editing the manuscript. BA directed the entire project and assisted with writing and editing the manuscript.
71
Abstract
In this Chapter, I describe application of a fluorescent reporter system that I developed to
systematically assess consequences of genetic perturbations on gene expression (Chapter 2).
I used the so-called Reporter-Synthetic Genetic Array (R-SGA) method to screen for
regulators of core histone gene expression. I discovered that the histone chaperone Rtt106
functions in a pathway with two other chaperones, Asf1 and the HIR complex, to create a
repressive chromatin structure at core histone promoters. Activation of histone (HTA1) gene
expression involves both relief of Rtt106-mediated repression by the activity of the histone
acetyltransferase Rtt109 and restriction of Rtt106 to the promoter region by the
bromodomain-containing protein Yta7. I propose that the maintenance of Asf1/HIR/Rtt106-
mediated repressive chromatin domains is the primary mechanism of cell cycle regulation of
histone promoters. These data suggest that the HIR/Rtt106 pathway may represent a
chromatin regulatory mechanism that is broadly used across the genome.
72
3.1 Introduction
Waves of cell cycle-regulated transcription are a universal feature of eukaryotic cell
cycles, yet our understanding of mechanisms linking gene expression to the cell cycle
remains incomplete. Conventional genetic screens in yeast, and complementary experiments
in many systems, have provided considerable insight but it is clear that many regulators
remain to be discovered. One important group of cell cycle-regulated genes encode histones
which form the nucleosome. Transcription of core histone genes is coordinated with the cell
cycle to ensure large amounts of new histones are available during DNA replication (Gunjan
et al., 2005). The restriction of histone gene expression to S phase (DNA synthesis) is
required not only to produce adequate histone pools but also to prevent toxicity that is
associated with their inappropriate overexpression at other stages of the cell cycle (Gunjan
and Verreault, 2003; Sopko et al., 2006). The yeast S. cerevisiae has proven a useful model
to understand the mechanism of regulation of core histone transcription. S. cerevisiae
contains two copies of each core histone gene, each of which is arranged in opposite
orientation to a gene encoding its dimer partner within the nucleosome: HHT1-HHF1 and
HHT2-HHF2, the two gene pairs that encode H3/H4, and HTA1-HTB1 and HTA2-HTB2, the
two gene pairs that encode H2A/H2B.
Four genes were identified in yeast genetic screens that encode transcriptional
repressors of three of the four histone gene pairs, HTA1-HTB1, HHT1-HHF1 and HHT2-
HHF2, both outside of S-phase and in response to hydroxyurea (HU), a chemical that causes
stalling of DNA replication forks (Osley and Lycan, 1987; Xu et al., 1992). These four
proteins, Hir1, Hir2, Hir3, and Hpc2, were subsequently demonstrated to co-purify as the
HIR protein complex (HIR) (Green et al., 2005; Prochasson et al., 2005). A fifth protein,
73
Asf1, also co-purifies with HIR (Green et al., 2005) and is similarly required to repress
transcription of HTA1-HTB1, HHT1-HHF1, and HHT2-HHF2 (Sutton et al., 2001). Asf1
and HIR are both histone chaperones, proteins that bind to histones and assemble or
disassemble chromatin (reviewed in De Koning et al., 2007). Asf1 and HIR are H3/H4-
specific chaperones that together are able to deposit histones onto DNA in a replication-
independent manner in vitro (Green et al., 2005).
Two members of the HIR protein complex, Hir1 and Hir2, share homology to the
HIRA protein in human cells (Lamour et al., 1995). Deletion of one copy of HIRA is
thought to underlie the human disease DiGeorge syndrome (Lamour et al., 1995). The N-
terminal region of HIRA shares homology to the yeast Hir1 protein, which contains a WD40
domain (Lamour et al., 1995). The C-terminal region of HIRA shares homology with the
yeast Hir2 protein suggesting that the human HIRA protein may be the result of a Hir1/Hir2
fusion (Lamour et al., 1995). The HIR protein complex in yeast has been largely
characterized in terms of its role in repressing transcription of histone genes to allow proper
S-phase specific expression of these genes.
The HIR proteins are also involved in heterochromatin silencing along with the CAF-
1 (chromatin assembly factor 1) complex, another H3-H4 histone chaperone. CAF-1 is a
complex of three proteins (Cac1, Cac2 and Msi1). Strains that harbour a mutation in both
HIR1 and members of CAF-1 show synergistic defects in heterochromatin silencing,
suggesting that HIR and CAF-1 may function in two separate pathways to create silenced
regions of chromatin (Sutton et al., 2001). Rtt106, another H3-H4 histone chaperone protein,
was found to interact physically with Cac1 and participate in heterochromatin silencing
(Huang et al., 2005). Subsequent studies revealed that the histone chaperones CAF-1 and
74
Rtt106 are required to recruit Sir1 and Sir3 to the silent mating type locus HMR in yeast
(Huang et al., 2007).
In addition to their role in heterochromatin silencing, Rtt106 and CAF-1 play a
distinct role from the HIR complex in replication-coupled nucleosome assembly where
Rtt106 and CAF-1 assemble nucleosomes onto newly replicated DNA during S-phase of the
cell cycle. Lysine 56 on histone H3 is acetylated (H3 K56Ac) by action of the histone
acetyltransferase Rtt109 and the associated histone chaperone Asf1, a chromatin mark that is
involved in genome stability and is also incorporated into the promoter regions of genes
(Collins et al., 2007; Driscoll et al., 2007; Han et al., 2007a; Masumoto et al., 2005; Recht et
al., 2006; Rufiange et al., 2007). Recent studies have shown that this acetylation mark
increases the affinity for Rtt106 and CAF-1 binding to histone H3 (Li et al., 2008). Although
Rtt106 and CAF-1 do not contain an acetyl lysine binding bromodomain, the only domain
previously known to bind acetylated histones (Kouzarides, 2007), Rtt106 was found to have
a PH domain similar to that in Pob3 (Li et al., 2008). Pob3 is a member of the FACT
complex and is involved in remodeling nucleosomes to allow RNA polymerase II elongation.
This PH domain in Rtt106 from residues 195 to 301 is necessary for binding K56Ac histone
H3 (Li et al., 2008). Further work showed that both Rtt106 and CAF-1 are required to
deposit H3 K56Ac histones onto newly replicated DNA (Li et al., 2008).
As noted above, repression of histone gene transcription has been characterized in
terms of the histone chaperone pathway that involves Asf1 and HIR. Asf1/HIR-mediated
repression of histone transcription relies on a specific DNA sequence, the negative regulatory
element (NEG), found in the promoters of the three HIR-regulated gene pairs but absent
from that of HTA2-HTB2 (Osley et al., 1986; Osley and Lycan, 1987). The NEG sequence is
75
required for proper cell cycle regulation of histone genes and when deleted, constitutive
histone transcription is observed (Osley, 1991; Osley et al., 1986). The only proteins known
to regulate HTA2-HTB2 are Spt10 and Spt21 (Dollard et al., 1994). Deletion of SPT10
reduces levels of all core histone gene transcripts to some degree (Hess et al., 2004; Xu et al.,
2005) possibly by acetylating histone H3 K56, although histone acetyltransferase activity for
Spt10 has not been shown. Additionally, the promoters of histone genes contain DNA
binding sites that are bound by Spt10 (Eriksson et al., 2005).
Although several regulators of histone gene expression are known, underlying
molecular mechanisms remain unclear. To address this void, I sought to exploit the
functional genomic tools available in yeast for performing rapid, saturating genetic screens.
Specifically, as described in Chapter 2, I devised a two-color GFP-RFP reporter system
called Reporter-Synthetic Genetic Array (R-SGA) to systematically assess the consequences
of gene deletions on a promoter of interest. I used R-SGA to screen an HTA1 reporter gene
and discovered a previously unappreciated role for the H3/H4 histone chaperone Rtt106 in
repression of histone gene expression. We demonstrate that Rtt106 functions downstream of
Asf1 and the HIR complex to create a repressive chromatin structure at the HTA1-HTB1
regulatory region. My genomic screen also revealed an activating role for Yta7, an
evolutionarily conserved protein containing both a bromodomain and an AAA ATPase
domain. Our molecular analysis indicates that Yta7 acts as a boundary element within the
HTA1 locus, preventing the spread of Rtt106 and associated repressive chromatin into the
histone gene coding regions. I also discovered that the HAT, Rtt109, functions as an HTA1
activator and genetic tests suggest that it counters HIR/Rtt106 repressive chromatin to permit
transcription in a cell cycle-regulated manner. Finally, our genome-wide analysis of
76
nucleosome positioning suggests that comparable HIR/Rtt106-Yta7 domains may dictate
regions of repressive chromatin throughout the genome.
3.2 Experimental Procedures
3.2.1 Yeast Strains and Plasmids
Yeast strains are listed in Table 3-1. Strains were constructed using standard yeast media
and genetic approaches. To generate the promoter-reporter construct, GFP (S65T) followed
by the ADH1 terminator sequence was amplified from the plasmid pFA6a-GFP S65T-ADH1
HIS3MX6 (Wach et al., 1997) and cloned into the SacI-PstI sites of plasmid pRS315
(Sikorski and Hieter, 1989), giving rise to the plasmid BA1926. The entire HTA1-HTB1
intergenic region was PCR amplified and cloned into plasmid BA1926 adjacent to GFP
leading to the reporter plasmid BA1930, which was subsequently transformed into the strain
BY4256 (Kainth et al., 2009).
77
Table 3-1: Strains used in this Chapter.
Strain Genotype Source BY4741 MATa, ura30, leu20, his31, met150 Winzeler et al., 1999 SC217 BY4741, asf1::KANMX Winzeler et al., 1999 SC218 BY4741, rtt109::KANMX Winzeler et al., 1999 JF08-405 BY4741, rtt106∆::KANMX Winzeler et al., 1999 JF08-406 BY4741, hir1∆::KANMX Winzeler et al., 1999 JF08-407 BY4741, yta7∆::KANMX Winzeler et al., 1999 6109 BY4741, swi4∆::KANMX EUROSCARF 1298 BY4741, spt10∆::KANMX EUROSCARF BY4742 MATα, his3Δ1, leu2Δ0, lys2Δ0 ura3Δ Winzeler et al., 1999 Y7092 MATα , can1Δ::STE2pr-his5, lyp1Δ Tong et al., 2007 BY4256 Y7092, HO::RPL39pr-tdTomato::hphMX Kainth et al., 2009 SN1601 Y7092, MATα, hir1∆::NAT Boone lab SN1483 Y7092, MATα, rtt106∆::NAT Boone lab BY4594 MATα, hir1∆::NAT, rtt109∆::KAN This work BY4595 MATα, rtt106∆::NAT, rtt109∆::KAN This work BY4596 MATα, rtt106∆::NAT, hir1∆::KAN This work JF08-32 BY4741, SPT10-TAP::HIS3 TAP fusion library JF08-25 BY4741, HIR1-TAP::HIS3 TAP fusion library JF08-402 BY4741, HIR2-TAP::HIS3 TAP fusion library JF08-403 BY4741, HIR3-TAP::HIS3 TAP fusion library JF08-404 BY4741, HPC2-TAP::HIS3 TAP fusion library JF08-21 BY4741, RTT106-TAP::HIS3 TAP fusion library JF08-000 BY4741, CAC1-TAP::HIS3 TAP fusion library JF08-26 BY4741, HIR1-TAP::HIS3, asf1∆::NAT This work JF08-294 BY4741, HIR1-TAP::HIS3, rtt106∆::KANMX This work JF08-293 BY4741, HIR1-TAP::HIS3, hpc2∆::KANMX This work JF08-22 BY4741, RTT106-TAP::HIS3, asf1∆::NAT This work JF08-23 BY4741, RTT106-TAP::HIS3, hir1∆::KANMX This work JF08-24 BY4741, RTT106-TAP::HIS3, rtt109∆::NAT This work JF08-256 BY4741, RTT106-TAP::HIS3, hir2∆::KANMX This work JF08-257 BY4741, RTT106-TAP::HIS3, hir3∆::KANMX This work JF08-258 BY4741, RTT106-TAP::HIS3, hpc2∆::KANMX This work JF08-255 BY4741, RTT106-TAP::HIS3, cac2∆::KANMX This work JF08-306 BY4741, SPT5-HA::KANMX This work JF08-278 BY4741, SPT4-TAP::HIS3 TAP fusion library JF08-282 BY4741, SPT6-TAP::HIS3 TAP fusion library JF08-270 BY4741, SPT16-TAP::HIS3 TAP fusion library JF08-285 BY4741, POB3-TAP::HIS3 TAP fusion library JF08-275 BY4741, YTA7-TAP::HIS3 TAP fusion library JF08-276 BY4741, YTA7-TAP::HIS3, hir1∆::KANMX This work JF08-277 BY4741, YTA7-TAP::HIS3, rtt109∆::KANMX This work JF08-295 BY4741, HIR1-TAP::HIS3, yta7∆::KANMX This work JF08-253 BY4741, RTT106-TAP::HIS3, yta7∆::KANMX This work BY4597 MATα, rtt106∆::KAN, yta7∆::NAT This work
78
3.2.2 SGA-based Functional Genomic Screen for Regulators of HTA1 expression
A detailed description of this method is described in Kainth et al., 2009 and in
Chapter 2. By using the SGA methodology (Tong, 2007), the HTA1pr-GFP reporter plasmid
(BA1930) along with the control RFP reporter gene was combined with each yeast deletion
mutant. Colony size for each arrayed mutant was derived using Qt ColonyImager software
version 1.01 (Boone and Andrews labs, unpublished) and positions on the array with no or
slow colony growth were removed from further analysis. Colony fluorescence was assayed
using the Typhoon Trio variable mode imager (GE Healthcare) and quantification was
carried out with GenePix Pro 3.0 software. Each screen was carried out in duplicate where
deletion mutants are represented twice on the array. GFP and RFP intensities were averaged
from replicate deletion mutant colonies on the array and log2 GFP:RFP ratios computed.
LOESS-normalized (Cleveland, 1979) log2 GFP:RFP ratios from duplicate screens were
averaged, giving rise to a single gene expression measurement for each deletion strain
derived from a total of 4 independent colonies. These log2 ratios were transformed to robust
Z-scores using median and median absolute deviation and P-values were assigned to these Z-
scores using the normal distribution. From a diverse dataset of 27 promoter-GFP reporter
screens, I obtained a list of 260 deletion mutants that usually appear as hits in these screens
(Andrews lab, unpublished data), which were removed from further analysis of specific
HTA1 promoter regulators.
3.2.3 qPCR Analysis of Histone Gene Expression
RNA was prepared using the RNeasy mini kit (Qiagen). The QuantiTect Reverse
Transcription Kit (Qiagen) was used to eliminate contaminating genomic DNA and to
79
synthesize cDNA from ~1 µg of RNA. qPCR reactions were set up using manufacturers
conditions with the TaqMan Core Reagents PCR kit (Applied Biosystems) containing
AmpliTaq Gold polymerase (Roche Molecular Systems Inc.) on a 7500 Real-Time PCR
block (Applied Biosystems). Primer annealing/extension for all qPCR reactions was 60 °C.
TaqMan gene specific probes were purchased from Applied Biosystems. To distinguish
between similar copies of histone genes, reverse primers annealing to the 3’ UTR of histone
transcripts were used.
3.2.4 Chromatin Immunoprecipitation (ChIP)
Soluble chromatin was prepared from cells treated with formaldehyde and
immunoprecipitated using standard procedures (Kim et al., 2004). Chromatin prepared from
TAP-tagged strains was incubated with IgG-Sepharose (Amersham). Antibodies against
histones H3 and H2B were obtained from Lake Placid Biologicals and used at a
concentration of 1:200. Immunoprecipitated DNA was analyzed by semi-quantitative,
multiplex PCR, always including an internal control (either a region of ACT1 or a non-
transcribed region of chromosome V, as indicated) for background. PCR reactions were
separated on 6% PAGE, imaged, and in some cases quantified by using a Typhoon
phosphoimager. For ChIP reactions that required quantification, the ratio of the experimental
to the control signal for the precipitated DNA was divided by the ratio of the experimental to
the control signal for the input DNA.
80
3.2.5 Purification and Analysis of Rtt106-associated proteins
Proteins associated with Rtt106-TAP were purified utilizing a modified TAP
purification procedure and subsequently identified by mass spectrometry. The mChIP
protocol consists of a single affinity purification step, whereby chromatin-bound protein
networks are isolated from mildly sonicated and gently clarified cellular extracts using
magnetic beads coated with antibodies (Lambert et al., 2009). This procedure reduces
sample loss due to poor
solubility of chromatin-associated protein complexes by shearing DNA through sonication to
maximize the protein complex solubility, and by reducing to a minimum the need for sample
centrifugation. The mChIP procedure was previously shown to be successful at purifying
proteins associated with both histone and non-histone chromatin-bound baits (Lambert et al.,
2009). Mass spectrometry analysis of gel slices was performed using LC-MS/MS as
described (Lambert et al., 2009).
3.2.6 Genome-wide nucleosome occupancy
Isolation of nucleosome bound DNA and hybridization onto the yeast tiling array
(Affymetrix) was carried out according to Lee et al., 2007.
3.3 Results
3.3.1 A dual-reporter functional genomic screen to discover new regulators of gene
expression
As described in Chapter 2, I devised a two-colour GFP-RFP reporter system called
Reporter-Synthetic Genetic Array (R-SGA) to systematically assess the consequences of
81
gene deletions on a promoter of interest (Figure 3-1A). The system involves creation of a
wild type yeast strain with the promoter sequence of a particular gene of interest fused to
GFP along with a control reporter construct with the constitutively expressed RPL39
promoter driving tdTomato (Shaner et al., 2004) (RFP) expression. The dual-reporter strain
is compatible with SGA methodology which enables marked genetic elements to be
combined in a single haploid cell through standard yeast mating and meiotic recombination
via an automated procedure (Tong et al., 2001; Tong et al., 2004). My goal was to survey
the yeast deletion collection (Giaever et al., 2002), which contains the set of ~4500 viable
KanMX-marked deletion mutants, for defects in gene expression. To do so, I apply the SGA
approach to introduce both the test and control fluorescent reporters into the deletion
collection. The resulting panel of yeast deletion mutants is then assayed for enhanced or
diminished promoter-GFP expression by scanning both fluorescence intensities directly from
colonies arrayed on agar plates using a scanning fluorimager. The ratio of GFP to RFP
fluorescence intensity for each yeast deletion mutant provides a genome-wide survey of the
effect of viable deletion mutants on the promoter of interest. I expect decreased GFP:RFP
when the deleted gene is a specific activator of the reporter gene, while deletion of a
repressor will result in higher GFP:RFP.
3.3.2 Identification of regulators of HTA1 expression
To uncover new regulators of S-phase specific expression of histone genes, I fused
the HTA1 promoter to GFP (HTA1pr-GFP) and used the R-SGA-based screening approach
described above (Figure 3-1A) to explore the genome for potential regulators of HTA1
expression. I carried out this screen in duplicate using a yeast deletion array where each
82
mutant is represented twice, providing an average GFP:RFP intensity from 2 deletion mutant
colonies for each replicate screen. Thus, log2 GFP:RFP ratios are averaged from 4 replicate
deletion mutants. As a test of significance of mutants causing differential GFP:RFP
expression, we assigned P-values to these log2 ratios based on the normal distribution of Z-
scores transformed from average log2 ratios from each screen. From replicate experiments,
we observed a Pearson correlation of 0.81.
83
Figure 3-1: Reporter-Synthetic Genetic Array (R-SGA) functional genomic screen for regulators of HTA1 expression. (A) Outline of R-SGA screening procedure describing construction of the output array having each deletion mutant combined with GFP and RFP reporter genes. Fluorescence is assayed directly from colonies arrayed on agar plates using a scanning fluorimager, and GFP:RFP ratios are calculated to assess specific effects of gene deletions on the GFP reporter of interest (see Methods). (B) Results of R-SGA screen for identification of regulators of the histone H2A gene HTA1. A reporter plasmid with the promoter normally driving HTA1 expression fused to GFP was screened as described in (A). Gene expression measurements taken from 3907 yeast deletion mutants are displayed, and mutants causing differential GFP expression with P-value < 10-4 are highlighted. (C) The fold change in GFP:RFP for each mutant is described along with corresponding P-values for each gene expression measurement.
84
Since the screen relies on GFP under the control of the HTA1 promoter, I expect
candidate regulators to reflect bona-fide promoter regulation rather than regulation of histone
mRNA stability, unless deletion mutants regulate the 5’ UTR of the transcript. As proof of
the utility of the approach, I identified a number of previously characterized HTA1 gene
regulators in my screen including HPC2, HIR1, HIR2, HIR3, ASF1 and SWI4 (Figure 3-1B)
(Green et al., 2005; Hess and Winston, 2005; Osley and Lycan, 1987; Prochasson et al.,
2005; Simon et al., 2001; Xu et al., 1992). Specifically, deletion of HPC2, HIR1, HIR3,
HIR2 or ASF1 caused HTA1pr-GFP expression to increase between 2 to 3.9-fold (Figure 3-
1C). Deletion of the known HTA1 transcriptional activator SWI4 caused a reduction in GFP
levels by 1.8-fold (Figure 3-1C). These results validate the utility of my approach and led
me to explore regulatory roles of other genes uncovered from my screen. These include the
histone H3-H4 chaperone, RTT106, which upon deletion caused a 2.5-fold increase in
HTA1pr-GFP expression, suggesting Rtt106 has a repressive role in HTA1 transcription. In
contrast, deletion of the H3-specific histone acetyltransferase, RTT109, caused a 2.5 fold
reduction in HTA1 expression (Figure 3-1C), which links Rtt109 to HTA1 activation. Other
genes uncovered by my screen that potentially encode activators of HTA1 expression include
YTA7, VPS75, and RRM3 (Figure 3-1C). Vps75 is a histone chaperone that copurifies with
the HAT Rtt109 and associates with Rtt109 to acetylate histone H3 K9 (H3 K9Ac)
(Fillingham et al., 2008). Rrm3 is a DNA helicase that helps replication forks pass protein-
DNA complexes (Ivessa et al., 2003). Yta7 encodes a bromodomain-containing protein that
was recently suggested to repress histone transcription (Gradolatto et al., 2008). However, in
my R-SGA screen, deletion of YTA7 resulted in decreased HTA1pr-GFP levels, indicating
Yta7 may have a more complex role at HTA1 than previously appreciated (see below).
85
To confirm my screen results, I used quantitative real time-PCR (qPCR) to assess
endogenous HTA1 transcript levels in deletion mutants of potential HTA1 regulators.
Consistent with the results of my genomic screen, deletion of RTT106 resulted in increased
HTA1 transcript levels (Figure 3-2A). The de-repression observed in an RTT106 deletion
strain was clear but slightly less pronounced than that observed in a strain lacking HIR1. In
my R-SGA screen, deletion of ASF1 caused de-repression of the promoter (Figure 3-1C) but
endogenous HTA1 transcript levels appear similar to wild type (Figure 3-2A) which was
observed previously and suggests a dual role in repression and activation of histone genes by
Asf1 (Sutton et al., 2001). One possibility is that Asf1 is behaving as an activating factor at
the HTA1 coding region which would be missed in my screen since the promoter is fused to
GFP. In terms of confirming activators uncovered from my R-SGA screen, deletion of YTA7
and RTT109 caused decreased HTA1 transcript levels relative to wild type, similar to the
deletion of the previously described histone activators SPT10 and SWI4 (Hess and Winston,
2005) (Figure 3-2A). Because deletion of SPT10 causes a significant growth defect, it is
absent on the deletion array and thus was not tested in my R-SGA screen.
The HIR proteins and Asf1 repress HTA1 expression outside of S-phase. To
determine if Rtt106 represses HTA1 in a manner similar to that of the HIR proteins, I
assessed HTA1 expression during the cell cycle in cells deleted for RTT106 and HIR1, which
have no obvious cell cycle defect by FACS analysis (data not shown). Wild-type, rtt106∆
and hir1∆ strains were arrested in late G1 phase with alpha factor and HTA1 transcripts were
profiled every 15 minutes using qPCR after release as a synchronized culture into fresh
medium. In wild type cells, HTA1 transcription fluctuated throughout the cell cycle, peaking
in S-phase before being repressed as cells progress past S-phase into G2/M (Figure 3-2B).
86
Consistent with previous results, deletion of HIR1 caused a significant defect in HTA1
expression, with obvious de-repression of gene expression throughout the cell cycle (Figure
3-2B). Similarly, the absence of RTT106 also caused constitutive HTA1 expression through
the cell cycle (Figure 3-2B). The HTA1 transcription peak occurred at 15-30 minutes after
alpha-factor release and was not completely repressed past S-phase of the cell cycle (45 and
60 minutes post release (Figure 3-2B)). Cell cycle regulation of CLB2 transcripts, which
peak at G2/M, was monitored to mark proper progression through the cell cycle. These
results indicate that the observed increase in HTA1 transcript levels in rtt106∆ log phase cells
is due to a failure to repress transcription outside of S-phase rather than over-activation
during S-phase of the cell cycle. It is important to note that cellular toxicity due to
inappropriate histone expression was not observed in these mutants, likely because
expression levels are not as high as those observed in other studies where constitutive histone
overexpression from GAL inducible histone genes affected cellular fitness (Gunjan and
Verreault, 2003; Sopko et al., 2006).
87
Figure 3-2: Rtt106, Rtt109 and Yta7 regulate histone gene expression. (A) cDNA was prepared from the indicated strains. The ratio of the indicated transcript to that of ACT1 was determined using qPCR. (B) Rtt106 represses HTA1 through the cell cycle. Each strain was blocked with 5µM -factor, released and samples taken at the indicated times. For each time point, cDNA was prepared and analyzed using qPCR. CLB2 transcription was assayed in addition to show proper progression through the cell cycle. (C) HIR-regulated histone genes are also regulated by Rtt106. Error bars represent standard deviations from the mean from at least 3 replicate qPCR reactions.
88
3.3.3 Rtt106 represses HIR-regulated histone genes
I next asked if Rtt106, like HIR, represses expression of other histone genes. First, I
tested the effect of RTT106 deletion on transcription of HTA1’s partner locus, HTB1. I did
not detect a significant effect of RTT106 deletion on transcription of HTB1 (Figure 3-2C),
possibly reflecting its demonstrated complex regulation by both transcriptional and post-
transcriptional mechanisms (Lycan et al., 1987; Xu et al., 1992). Both Asf1 and HIR repress
HHT1, HHF1, HHT2 and HHF2 gene expression. Likewise, I saw that deletion of RTT106
resulted in higher levels of HHT1, HHF1, HHT2 and HHF2 transcripts (Figure 3-2C),
indicating that the HHT1-HHF1 and HHT2-HHF2 loci are also regulated by Rtt106. Unlike
the other histone loci, HTA2-HTB2 is not subject to Asf1/HIR repression. Similarly, I found
that Rtt106 did not repress transcription of HTA2 and HTB2 (Figure 3-2C). I conclude that
Rtt106, like HIR and Asf1, represses transcription at three of the four histone gene pairs.
Rtt106 localizes to HTA1-HTB1 in a manner dependent on Asf1 and the HIR complex
My data link the histone chaperone Rtt106 to repression of HTA1-HTB1 expression.
To ask if Rtt106 acts directly on the HTA1-HTB1 promoter, we used chromatin
immunoprecipitation (ChIP) to assess whether Rtt106 and the four members of the HIR
complex localize to the HTA1-HTB1 region. Figure 3-3A shows the HTA1-HTB1 locus with
approximate locations of primer sets used in our ChIP analysis. As a control, we found that
Spt10-TAP cross-linked most effectively to the promoter region containing the NEG site as
well as several upstream activating sequences (primer set ‘C’, Figure 3-3B), consistent with
previous results (Eriksson et al., 2005; Xu et al., 2005).
89
We next used ChIP to assay the cross-linking pattern of Rtt106 and the four members
of the HIR complex at HTA1-HTB1. Hir2 is known to cross-link at the promoter region of
HTA1-HTB1 (Green et al., 2005). Consistent with this result, we found that all four members
of the HIR complex, Hir1-TAP, Hir2-TAP, Hir3-TAP and Hpc2-TAP, specifically localized
to this region (Figure 3-3B and data not shown). HIR binding was restricted to only
background levels outside of region ‘C’ and at the ORFs (Figure 3B and data not shown).
Importantly, like the HIR proteins, Rtt106-TAP specifically localized to the HTA1-HTB1
promoter region and did not localize to the ORFs (Figure 3-3B).
90
Figure 3-3: Rtt106 and HIR localize to the promoter region of HTA1-HTB1. (A) Schematic representation of the PCR products (A-E) used in ChIP to cover the HTA1-HTB1 locus. (B) Rtt106 and members of the HIR complex (but not Cac1) cross-link to the promoter region of HTA1-HTB1. PCR results from IgG-sepharose ChIP of an untagged negative control (WT) or positive control (Spt10-TAP) show that PCR products A-E effectively cover the locus. Precipitated chromatin was used for PCR amplification (upper panels). The top band is specific to the HTA1-HTB1 locus, while the common lower band (marked by an asterisk) is an internal background control from a nontranscribed region on chromosome V. The bottom panels show the input control. (C) Rtt106 and members of HIR cross-link to the promoters of the same set of histone genes. A similar analysis was performed as in Figure 3-3A with primers directed against the promoters of the indicated histone genes. ChIP analysis was performed exactly as in Figure 3-3B.
91
Since HIR and Rtt106 repress three of the four histone loci (Figure 3-2C), we tested
whether they also localize to other histone promoters to further correlate transcriptional
effects with promoter binding. We used ChIP of Spt10-TAP to the promoter regions of the
four core histone promoters as a control for the performance of our primers. We also
assessed binding to the promoter region of the HTZ1 gene, which encodes an H2A variant
and serves here as a negative control since its transcription is not under cell cycle control. As
expected (Eriksson et al., 2005), Spt10-TAP effectively localized to the promoter regions of
the four core histone promoters, but not to that of HTZ1 (Figure 3-3C). Hir1-TAP also
cross-linked to the promoters of HTA1-HTB1, HHT1-HHF1 and HHT2-HHF2 but not HTA2-
HTB2 (Figure 3-3C), consistent with the failure of the HIR complex to regulate HTA2-
HTB2. Like Hir1, Rtt106-TAP also localized to the promoter region of HTA1-HTB1 and was
enriched at HHT1-HHF1 and HHT2-HHF2, although not to the same degree as at HTA1-
HTB1 (Figure 3-3C). Like Hir1, Rtt106 did not cross-link above background levels to
HTA2-HTB2. Thus, promoter localization of Hir1 and Rtt106 correlates with their ability to
specifically repress transcription of HTA1-HTB1, HHT1-HHF1 and HHT2-HHF2 but not
HTA2-HTB2.
So far, our experiments place HIR and Rtt106 in a common pathway that functions to
repress histone transcription. To further explore the relationship between Rtt106 and other
histone gene regulators, we used our ChIP assay to test the genetic requirements for specific
cross-linking of Hir1 and Rtt106 to the promoter of HTA1-HTB1. Consistent with previous
results (Green et al., 2005), deletion of HPC2 but not ASF1 prevented recruitment of Hir1-
TAP to the HTA1-HTB1 promoter (Figure 3-4A). Similar to ASF1, deletion of RTT106 did
not affect Hir1-TAP recruitment to HTA1-HTB1 (Figure 3-4A). However, when we deleted
92
either HIR1 or ASF1, recruitment of Rtt106 to HTA1-HTB1 was undetectable (Figure 3-4B).
Rtt106-TAP recruitment to the HTA1-HTB1 promoter was also prevented by deletion of
HIR2, HIR3, or HPC2 (Figure 3-4C).
Asf1 and HIR form a protein complex in yeast (Green et al., 2005) and the interaction
of Asf1 with the HIR complex is abolished when any of the four HIR subunits are deleted.
These results suggest that an intact Asf1-HIR complex functions upstream of Rtt106
recruitment to HTA1-HTB1. By contrast, although Rtt106 physically interacts with the CAF-
1 protein complex (Huang et al., 2005), localization of Rtt106 to HTA1-HTB1 occurred
independently of Cac2, a subunit of the CAF-1 complex (Figure 3-4C) indicating that
Rtt106 is functioning in a CAF-1 independent pathway with the Asf1/HIR complex. This
result is consistent with the failure of Cac1-TAP to cross-link to the region (Figure 3-3B).
In order to determine the molecular basis of the HIR requirement for Rtt106 recruitment, we
co-purified its associated proteins and identified them using mass spectrometry (Lambert et
al., 2009). We co-purified two of the three members of the CAF-1 complex (Figure 3-4D)
along with Pol30/PCNA (which interacts with CAF-1), consistent with previous results
(Huang et al., 2005). In addition, we co-purified three of the four members of the HIR
complex (Figure 3-4D). Our genetic experiments demonstrating Asf1/HIR-dependent
localization of Rtt106 to HTA1-HTB1, coupled with these biochemical data, suggest a direct
physical interaction between these proteins at a core histone promoter.
93
Figure 3-4: HIR and Asf1 are required for Rtt106 localization to HTA1-HTB1. (A) Hir1 localization to the HTA1-HTB1 promoter is independent of Asf1 and Rtt106. The analysis is the same as in Figure 3-3 with the exception that the top band represents PCR product ‘C’ from HTA1-HTB1. (B) Rtt106 localization to the HTA1-HTB1 promoter requires Asf1 and Hir1. The analysis is the same as above except that the internal background control (indicated by an asterisk) is from the ACT1 gene. Below the ChIP analysis, a western blot indicates that RTT106 expression is not dependent on Asf1 or Hir1. (C) Rtt106 localization to the HTA1-HTB1 promoter requires all members of HIR but not CAF-1. ChIP analysis is the same as in Figure 3-4A. Below the ChIP analysis, a western blot indicates that RTT106 expression is not dependent on expression of HIR members. (D) Affinity purification and identification of Rtt106-TAP associated proteins. A silver-stained SDS-PAGE is shown with affinity purified proteins from an untagged strain (-) and from Rtt106-TAP (+). Co-purifying proteins were identified by LC-MS/MS as described in Lambert et al., 2009. The percent sequence coverage is indicated in the table, with the number of unique peptides shown in parentheses.
94
3.3.4 The HTA1-HTB1 promoter region is nucleosome-free in asf1, hir1, rtt106 mutants
As described in the Introduction, Asf1, the HIR complex and Rtt106 are histone
chaperones/chromatin assembly factors. To ask if their effects on HTA1-HTB1 transcription
are related to this function, we used an antibody generated against unmodified histone H3 to
assess H3 levels at the HTA1-HTB1 promoter in WT and several deletion strains (Figure 3-
5A). Compared to a WT strain, or a strain deleted for RTT109 which we also identified in
our screen (see below), the amount of H3 that cross-linked to the HTA1-HTB1 promoter in
asf1, hir1, and rtt106 was low relative to a control locus (Figure 3-5A). A similar
experiment with an antibody against unmodified histone H2B (Figure 3-5B) revealed that
histone H2B levels were significantly lower again in asf1, hir1, and rtt106 mutants
relative to a WT strain (Figure 3-5B).
95
96
Figure 3-5: Asf1, HIR and Rtt106 collaborate to assemble chromatin at the HTA1-HTB1 promoter. (A, B) Chromatin was prepared from the indicated strains and immunoprecipitated with an antibody against histone H3 (A) or H2B (B). The top band in the ChIP analysis represents PCR product ‘C’ from HTA1-HTB1 and the internal background control (indicated by an asterisk) is from the ACT1 gene. (C, D) A genome-wide nucleosome positioning assay was used to identify regions of depleted nucleosomes in rtt106Δ and hir1Δ strains. The HTA1-HTB1 intergenic region is nucleosome free when RTT106 or HIR1 is deleted but has regions of higher and lower intrinsic occupancy in the in vitro experiments (C). In Panel (D) nucleosome profiles at promoter regions genome-wide are sorted in ascending order (most depleted in blue colour to most occupied in yellow colour) based on the average nucleosome occupancy up to 500 base pairs upstream of the transcriptional start site (TSS) for each ORF. The 50 top ranking nucleosome depleted promoter regions are shown. Histone genes are highlighted in red while other nucleosome depleted regions overlapping in rtt106Δ and hir1Δ strains are highlighted in cyan.
97
To further explore the molecular defect in histone chaperone mutants, we assessed
genome-wide nucleosome occupancy of promoter regions in wild type cells as well as hir1∆
and rtt106∆ strains using a previously described method (Lee et al., 2007). Consistent with
our ChIP results, we found that most of the HTA1-HTB1 intergenic region was nucleosome-
free when either HIR1 or RTT106 was deleted while nucleosome positions at HTA1-HTB1
coding regions remained unchanged (Figure 3-5C). Similarly, we found that along with
HTA1-HTB1, promoter regions of HHT1-HHF1 and HHT2-HHF2 were amongst the most
nucleosome-depleted intergenic regions genome-wide in strains deleted for either HIR1 or
RTT106 (Figure 3-5D, red). These results suggest that cell-cycle repression of HTA1-HTB1,
HHT1-HHF1 and HHT2-HHF2 is dependent on a nucleosome assembly pathway that relies
on the coordinated actions of Asf1, the HIR complex, and Rtt106. A number of other
promoters in the genome are also nucleosome-free in the HIR1 and RTT106 deletion strains,
including several common to both (Figure 3-5D, cyan), suggesting HIR and Rtt106 may
function together at other loci.
Since the HTA1-HTB1 promoter region is nucleosome-free in the rtt106Δ and hir1Δ
mutants, we wondered whether this promoters intrinsic DNA sequence preference favours
nucleosome depletion. This would indicate that the action of HIR and Rtt106 are required
for proper nucleosome occupancy at this promoter. To investigate this, we assessed the
DNA-encoded nucleosome occupancy at the HTA1-HTB1 promoter using two separate in
vitro nucleosome occupancy experiments generated by Kaplan et al. (2009) and Zhang et al.
(2009). In these experiments, histone octamers were incubated with yeast genomic DNA and
the nucleosome occupancy, determined solely by DNA sequence, was determined as
described in Section 1.4 (Kaplan et al., 2009; Zhang et al., 2009). In Figure 3-5C, the in
98
vitro occupancy data is shown at the HTA1-HTB1 promoter and reveals that in the central
region (which encompasses the NEG site) of the promoter, there appears to be higher
intrinsic nucleosome occupancy relative to the TSS. This indicates that in wild type cells,
nucleosome occupancy is determined by the action of HIR, Rtt106 and to some degree the
DNA sequence in the region of the NEG site. When HIR1 or RTT106 are deleted, another
factor may be disassembling nucleosomes at the promoter that form because of preference
for the DNA sequence itself, which means there may be constant assembly/disassembly
occurring at the HTA1-HTB1 intergenic region. It should be noted that discrepancies exist in
the in vitro nucleosome occupancy data at the HTA1-HTB1 locus in these two data sets
(particularly at the HTB1 TSS), possibly due to differences in experimental and analysis
methods (Kaplan et al., 2009; Zhang et al., 2009).
3.3.5 HIR/RTT106 repression at HTA1-HTB1 creates a requirement for RTT109
Our analysis of HIR and RTT106 requirements for histone gene expression and
promoter binding suggest that activation of HTA1 may reflect relief of repression, rather than
the function of specific activators, as seen at other cell cycle-regulated promoters (reviewed
in Wittenberg and Reed, 2005). To test this idea further, we first used a synchronized cell
culture and our ChIP assay to assess Rtt106 localization at the promoter of HTA1-HTB1
throughout the cell cycle. We found that the proportion of HTA1-HTB1 promoter that bound
Rtt106 did not change significantly during the cell cycle (Figure 3-6A). In addition,
consistent with our previous results, Rtt106 did not cross-link to the ORF region of HTA1 at
any time in the cell cycle (Figure 3-6A). qPCR analysis of HTA1 expression confirmed that
the cells progressed synchronously through the cell cycle in our experiment (Figure 3-6A).
99
Thus, temporal changes in Rtt106 localization to the promoter cannot account for repression
of cell cycle transcription of the histone genes. Rather, we reasoned that an activating factor
may serve to relieve Rtt106-mediated repression during S-phase. As noted earlier, our
functional genomic screen uncovered RTT109 as an activator of HTA1 expression (Figure 3-
1B and 3-2A). The Rtt109 histone acetyltransferase has specificity for K56 on Histone H3
(H3 K56ac) (Collins et al., 2007; Driscoll et al., 2007; Han et al., 2007a) and acetylation of
H3K56 is required for the cell-cycle dependent transcription of the histone genes (Xu et al.,
2005). I therefore examined HTA1 expression in a series of double mutants by qPCR. As
expected, HTA1 expression was de-repressed in hir1 and rtt106 single mutants (Figure 3-
6B). HTA1 expression was not considerably different in a hir1 rtt106 double mutant than
the respective single mutants (Figure 3-6B), consistent with them functioning together in the
same pathway. As I showed previously, rtt109∆ caused a reduction in HTA1 transcript levels
(Figures 3-2A and 3-6B). However when either hir1 or rtt106 was combined with
rtt109, the inhibitory effect of rtt109 on HTA1 expression was significantly reduced
(Figure 3-6B). These genetic results suggest that Rtt109 functions to activate histone gene
expression by antagonizing the repressive effect of Rtt106 and the HIR complex.
100
Figure 3-6: Constitutive repression at HTA1-HTB1 creates a requirement for RTT109. (A) A Rtt106-TAP strain was blocked with -factor, released and samples taken for ChIP and qPCR at the indicated times. The top band of the ChIP analysis represents PCR product ‘C’ from HTA1-HTB1 while the common lower one (marked by an asterisk) is an internal background control from a nontranscribed region on chromosome V. The top left panel represents PCR products from Rtt106-TAP ChIP DNA at region “C” of the HTA1-HTB1 locus while the top right panel represents PCR products from Rtt106-TAP ChIP DNA at region “E” of the HTA1-HTB1 locus. qPCR shows normal kinetics of HTA1 expression throughout the cell cycle. (B) The cDNA was prepared from the indicated strains. The ratio of the indicated transcript to that of ACT1 was determined using qPCR. Error bars represent standard deviations from the mean from at least 3 replicate qPCR reactions.
101
3.3.6 Yta7 is a boundary element within the HTA1-HTB1 locus
Because Rtt106 and members of the HIR complex have recently been identified as
transcriptional elongation factors (Imbeault et al., 2008; Nourani et al., 2006), we compared
their localization pattern at HTA1-HTB1 to that of several other elongation factors implicated
in chromatin assembly. We used ChIP to assess the localization of the functionally related
proteins Spt4-TAP, Spt5-TAP and Spt6-TAP, as well as the two subunits of FACT, Pob3-
TAP and Spt16-TAP, at the HTA1 portion of the HTA1-HTB1 region. All five factors cross-
linked at HTA1 in a pattern distinct from HIR/Rtt106. While HIR/Rtt106 cross-linked
primarily to the promoter region of HTA1 (Figure 3-3B and data not shown), Spt4, Spt5,
Spt6, Spt16 and Pob3 associated mainly with the coding regions of HTA1 (Figure 3-7) and
HTB1 (data not shown). Thus, consistent with our other experiments, HIR-Rtt106 likely has
a role at the HTA1 locus distinct from the elongation factors that we tested.
102
Figure 3-7: Spt4, Spt5, Spt6 and FACT cross-link to the transcribed regions of HTA1 but not to the promoter region. The top band of the ChIP analysis represents the indicated PCR product from HTA1-HTB1 while the common lower band (marked by an asterisk) is an internal background control from a nontranscribed region on chromosome V.
103
We next assayed the relationship between YTA7 and the various chromatin assembly
factors and chaperones that are present at histone core promoters. As noted earlier, YTA7
encodes a bromodomain-containing protein and was identified as an HTA1 activator in our
R-SGA screen (Figure 3-1B). Yta7 is known to function as a boundary element within
chromatin at HMR, a locus that is typically transcriptionally silent (Tackett et al., 2005). To
ask if Yta7 might perform a similar function at a transcribed locus, we assayed cross-linking
of Yta7 at HTA1-HTB1. We observed strong enrichment of Yta7 at regions occupied by
HIR/Rtt106 (region “C”), Spt4/5/6 and FACT (region “E”) and also to the region between
them (region “D”) (Figure 3-8A). Similar to the Hir1 and Rtt106 proteins, we found Yta7
cross-linked to the promoter region of HHT1-HHF1 and HHT2-HHF2 but not HTA2-HTB2
(Figure 3-8B). Localization of Yta7 was dependent on HIR1 but not RTT109 (Figure 3-8A)
at all three histone promoter regions (Figure 3-8B). Because Yta7 localized efficiently to
the region bounded by HIR/Rtt106 and Spt4, Spt5, Spt6, and FACT, we asked whether HIR
or Rtt106 localization to HTA1-HTB1 was affected by deletion of YTA7. We did not observe
a difference in the cross-linking pattern of Hir1-TAP at HTA1-HTB1 in a yta7∆ strain
(Figure 3-8C).
104
Figure 3-8: Yta7 localizes to the HTA1-HTB1 locus. (A) Hir1 is required for Yta7 localization to HTA1-HTB1. The top band of the ChIP analysis represents the indicated PCR product from HTA1-HTB1 while the common lower band (marked by an asterisk) is an internal background control from a nontranscribed region on chromosome V. A western blot shows that YTA7 expression is not dependent on expression of Hir1. (B) Yta7 localization to other histone promoters is dependent on HIR1. A similar analysis was performed as in Figure 3-8A with primers directed against the promoters of the indicated histone genes. (C) Yta7 is not required for proper Hir1 localization at HTA1-HTB1. ChIP analysis was performed as in Figure 3-8A.
105
Instead, we found that Rtt106 cross-linked throughout the entire HTA1-HTB1 region in the
absence of YTA7, including the transcribed regions where it is normally not present (Figure
3-9A). Thus Yta7 influences Rtt106 but not Hir1 localization at HTA1-HTB1. We propose
that the inhibitory affect that deletion of YTA7 causes on HTA1 transcript levels (Figure 3-
9B) is the result of Rtt106 mislocalization at the HTA1 coding region, likely by creating
repressed chromatin or inhibiting transcription initiation by RNA polymerase II. When the
YTA7 deletion is combined with deletion of RTT106, the inhibitory effect on HTA1
transcription is partially relieved (Figure 3-9B), suggesting Rtt106 mislocalization in the
absence of Yta7 causes a defect in HTA1 expression. We also saw increased Rtt106 cross-
linking to the HHT1-HHF1 and HHT2-HHF2 promoters in the absence of YTA7, indicating
the same relationships among Hir1, Rtt106 and Yta7 at three of the four histone gene pairs
(Figure 3-9C). We conclude that Yta7 may contribute to the proper activation of histone
gene expression by preventing Rtt106 from spreading from the HTA1-HTB1 regulatory
region into the transcribed regions (Figure 3-10).
106
Figure 3-9: Yta7 creates a boundary within the HTA1-HTB1 locus. (A) Yta7 is required for proper Rtt106 localization at HTA1-HTB1. The top band of the ChIP analysis represents the indicated PCR product from HTA1-HTB1 while the common lower band (marked by an asterisk) is an internal background control from a nontranscribed region on chromosome V. (B) qPCR analysis of HTA1 transcript levels reveals that the inhibitory affect caused by deletion of YTA7 can be partially relieved in the yta7Δ rtt106Δ strain. (C) Deletion of YTA7 affects localization of Rtt106-TAP to other histone promoters. Rtt106-Tap localizes to other histone promoters with the exception of HTA2 and in the absence of YTA7, more Rtt106-TAP cross links to these promoters indicating that Yta7 functions as a boundary element at HIR/Asf1/Rtt106 regulated histone promoters. ChIP analysis was performed as described in Figure 3-9A with primers directed against the promoters of the indicated histone genes.
107
3.4 Discussion
Cell cycle-dependent regulation of histone gene expression is a universal feature of
eukaryotic cell cycles, yet the mechanisms of activation and repression of histone genes have
remained obscure. I developed a reporter-based functional genomic screen using SGA
technology in S. cerevisiae and identified several regulators of HTA1 expression,
demonstrating that the H3-H4 histone chaperone Rtt106 functions along with known HTA1
regulators, Asf1 and the HIR complex, to repress HTA1 transcription. We found that the
HTA1-HTB1 promoter region is mostly nucleosome-free in the absence of ASF1, HIR1 or
RTT106. Whether or not Asf1/HIR/Rtt106-mediated nucleosome formation is associated
with a specific histone post-translational modification pattern remains unknown. We also
discovered that Yta7 bound the HTA1-HTB1 promoter region and that loss of Yta7 results in
a defect in activation of HTA1 transcription. Yta7 appears to mediate histone gene activation
by restraining repressive chromatin formed by Asf1/HIR/Rtt106 at the HTA1-HTB1
promoter.
Our genetic and biochemical experiments suggest that a primary role of the HIR
complex and Asf1 in histone gene regulation is to recruit the histone H3/H4 chaperone
Rtt106 to promoter regions. We found that Rtt106 is present at the HTA1-HTB1 regulatory
region throughout the cell cycle (Figure 3-6A), suggesting that repression is the default state
at HTA1-HTB1. I propose that, unlike other well characterized cell cycle-sensitive
promoters, activation of histone gene expression does not require the action of specific
activating transcription factors, although they may play some role. Rather, the key to
activating histone gene expression resides with overcoming the repressive chromatin
structure established by Rtt106 and its partners. My functional genomic screen and follow-
108
up experiments suggest that Rtt109 may relieve repression by Asf1/HIR/Rtt106 at HTA1-
HTB1 (Figure 3-10).
109
Figure 3-10: A model describing histone chaperone mediated repression at the HTA1 locus in yeast. See text for details.
110
The defect in HTA1 activation caused by mutation of RTT109 is partially overcome by
deletion of hir1∆ or rtt106∆. The remaining repressive effect of the RTT109 deletion when
combined with hir1 or rtt106 could be a consequence of its requirement to remove any
remaining nucleosomes in those double mutants. Thus, the activity of Rtt109 may not be
required when the HTA1-HTB1 promoter region is nucleosome-free. This phenomenon
mirrors that of the HIR-dependent recruitment of the yeast SWI/SNF complex that also
activates transcription at HTA1-HTB1 (Dimova et al., 1999) since mutation of components of
the HIR complex abolishes the requirement for SWI/SNF in transcriptional activation.
Rtt109 acetylates H3 K56, a modification that is enriched at the yeast histone gene
promoters (Xu et al., 2005) in a cell cycle-dependent manner. The acetylation of H3 K56 is
required for the recruitment of the SWI/SNF complex member Snf5 (Xu et al., 2005). Thus
SWI/SNF could act directly downstream of Rtt109 in a cell cycle-dependent manner to
overcome HIR/Rtt106-mediated repression and activate transcription of HTA1. Rtt109 and
H3 K56Ac have been implicated in the process of nucleosome disassembly leading to
transcriptional activation at the PHO5 locus (Williams et al., 2008). Based on these
observations, a plausible model for Rtt109 action at the HTA1-HTB1 promoter involves the
coordinated action of Rtt109 and SWI/SNF to disassemble nucleosomes leading to
transcriptional activation (Figure 3-10). SWI/SNF functions in nucleosome eviction
pathways [for example, at SUC2 where it binds to the UAS and mediates nucleosome
eviction (Schwabish and Struhl, 2007)].
It is likely that Rtt109 is not itself recruited to the histone promoters since it only
acetylates non-nucleosomal histones (Han et al., 2007b) and we were unable to detect
Rtt109-TAP at the HTA1-HTB1 locus using ChIP (data not shown). One possibility is that
111
during S-phase, Rtt109 acetylates H3 K56 on free histone H3 which is then incorporated into
the HTA1-HTB1 intergenic region because of the presence of Asf1/HIR/Rtt106 since Rtt106
has been shown to have a higher affinity for K56 acetylated histone H3 (Li et al., 2008).
HIR is directed to the histone promoter because of the presence of the NEG cis-regulatory
site which is required for HIR localization (Osley, 1991). Deletion of the NEG sequence or
HIR results in constitutive histone transcription (Osley, 1991; Osley et al., 1986; Osley and
Lycan, 1987) consistent with our results showing that deletion of Rtt106 also causes de-
repression of HTA1 transcription and that localization of Rtt106 to the NEG region is
dependent on HIR. Outside of S-phase, removal of H3 K56Ac by the Hst3/4 deacetylase
(Celic et al., 2006; Maas et al., 2006) would result in incorporation of unacetylated H3 K56
histones by the Asf1/HIR/Rtt106 complex [which is likely constitutively present at the
promoter (Figure 3-6A)] leading to repression of histone gene expression. The presence of
Yta7 at the promoter may be dependent on other histone marks in the histone N-terminal
domain since previous work has shown that Yta7 does not associate with K56Ac histone H3
(Gradolatto et al., 2008). However, it is likely that Yta7 localization specifically at the
HTA1-HTB1 region is dependent on the presence of NEG, since this site is required for HIR
recruitment (Osley, 1991) which is in turn required for Yta7 localization (Figure 3-8A and
B). An important caveat to these models is the unclear relationship of Rtt109 to Spt10, a
protein initially suggested to be the H3 K56-specific HAT at the histone genes in S.
cerevisiae (Xu et al., 2005). My results indicate that both Rtt109 and Spt10 function to
activate HTA1 (Figure 3-2). Clearly more work will be required to discover whether Spt10
directly acetylates H3 K56, or if it stabilizes H3 K56Ac-containing nucleosomes, as well as
what (if any) relationship exists between Spt10 and Rtt109 at HTA1-HTB1. Additionally, we
112
also detected another histone chaperone, Vps75, as a potential activator of HTA1
transcription (Figure 3-1B). Vps75 physically interacts with Rtt109 which acetylates H3 K9
(Fillingham et al., 2008). However, we have not observed a significant change in HTA1
transcription in a point mutant where H3 K9 cannot be acetylated (H3 K9R mutant), nor have
we seen Vps75 localization at the HTA1-HTB1 locus, indicating that H3 K9 acetylation is
likely not required for HTA1 activation (data not shown). Vps75 is required for stability of
Rtt109 (Fillingham et al., 2008) meaning that Vps75 may result in lower Rtt109 levels and
thus a slight reduction in HTA1 transcription, consistent with our screening results.
We also identified the DNA helicase RRM3 as a potential activator of HTA1
transcription (Figure 3-1B). Interestingly, a screen for proteins required to localize
telomeres to the nuclear periphery revealed a role for Rtt109, Asf1, Vps75, H3 K56Ac and
Rrm3 (Hiraga et al., 2008), indicating these proteins along with K56Ac might function
together in a common pathway that might also be applicable to transcription of the HTA1
gene. Clearly future work will be required to determine what role, if any, RRM3 plays in
histone gene transcription. In follow-up work, ChIP analysis will be carried out to determine
if RRM3 localizes to the HTA1-HTB1 locus.
Rtt106 interacts physically with CAF-1 (Huang et al., 2005) to function in
replication-coupled chromatin assembly (Li et al., 2008). In contrast, both the yeast
Asf1/HIR complex and higher organism versions of HIR function in the replication-
independent assembly of chromatin (Green et al., 2005 and reviewed in De Koning et al.,
2007). We found that the CAF-1 subunit Cac2 is not necessary for Rtt106 localization to
HTA1-HTB1 (Figure 3-4), suggesting that Asf1/HIR/Rtt106-mediated nucleosome assembly
at HTA1-HTB1 likely occurs in a replication-independent manner. Although our results
113
suggest that the major function of the three histone chaperones at the HTA1-HTB1 promoter
is the assembly of nucleosomes, the precise nature of Asf1/HIR/Rtt106-mediated repression
remains to be determined.
A striking feature of the spectrum of regulatory proteins at the HTA1-HTB1 locus is
the physical separation of HIR/Rtt106 from Spt4, Spt5, Spt6 and FACT (Figures 3-3B and
3-7). Rtt106 was recently found to cross-link to the coding region of PMA1 (Imbeault et al.,
2008), as do Spt4, Spt5, Spt6 and FACT (Kim et al., 2004). Spt6 and Rtt106 are also known
to function in parallel to suppress cryptic initiation at an internal promoter within the FLO8
gene (Imbeault et al., 2008). Since Rtt106 has a defined role in transcriptional elongation, a
mechanism may exist to restrict Rtt106 to the promoter region of HTA1-HTB1. Yta7 was
originally identified as a protein whose absence led to the spreading of the silent state of
chromatin at HMR to surrounding genes, consistent with its proposed function as a barrier
between regions of heterochromatin and euchromatin at HMR (Tackett et al., 2005). We
found that Rtt106, but not Hir1, cross-links throughout HTA1-HTB1 in the absence of YTA7,
including the transcribed regions (Figure 3-8C and Figure 3-9A). Thus, the loss of a barrier
protein (Yta7) at HTA1-HTB1 appears to cause the lateral spread of Rtt106 from the
promoter through the ORFs. Since Rtt106 is a histone chaperone specifically associated with
the formation of repressive chromatin, its lateral spreading across the coding region of HTA1
could repress transcription by propagating a repressive chromatin structure. The regulation
of Rtt106 localization could represent a more general mechanism underlying
heterochromatin spreading. For example Yta7 may influence Rtt106 at heterochromatin
boundaries at HMR, a region where both proteins have been functionally implicated (Huang
et al., 2007; Jambunathan et al., 2005; Tackett et al., 2005). More generally, Yta7 could
114
function together with Asf1/HIR/Rtt106 to create short domains of repressed chromatin
throughout the genome. Our nucleosome occupancy data hints that the promoter regions of
other genes are indeed subject to HIR/Rtt106 regulation and Yta7 may also be involved in
their regulation. Global approaches such as ChIP-chip combined with nucleosome
occupancy studies will address this question and are currently in progress.
In this chapter, I present a detailed analysis of histone gene regulation based on the
use of a dual-reporter screen to discover new regulators. I have elucidated mechanisms of
histone gene control for both well-studied regulators like HIR and previously unknown
proteins like the histone chaperone Rtt106 and the Yta7 boundary element. I also note that
the functional genomics approach presented here can be applied to study virtually any
pathway for which an appropriate fluorescent reporter gene can be devised, providing a
powerful means to link gene function to transcriptional control.
115
Chapter 4
Summary and Future Directions
116
4.1 Summary
Reporter gene screens are used as a fundamental tool to identify regulatory proteins
that control gene expression from promoters of interest. In the past, forward genetic screens
were carried out to randomly mutagenize cells harbouring reporter constructs with the aim of
mapping mutant genes causing differential reporter expression. This general approach has
identified many transcriptional regulators but has several limitations. Screens involving
random mutagenesis can be quite time consuming, since they require generating a large
library of mutants to attempt an effective survey of the genome, and follow-up mapping of
mutations causing differential reporter expression requires considerable work. Also, these
screens are generally not quantitative and are rarely saturating.
Systematic, genome-wide analysis of mutants for an effect on reporter gene
expression in a quantitative, rapid and unbiased manner was not possible, and my thesis work
aimed to address this. I describe the development and application of a screening system that
combines fluorescent proteins with functional genomic tools and resources already available
in budding yeast. Specifically, I developed a two-colour reporter system where a wild type
yeast strain that harbours a promoter of interest fused to GFP along with a control promoter
fused to RFP can be introduced into an ordered array of yeast deletion mutants using the
SGA methodology. GFP and RFP fluorescence are easily assayed by directly scanning the
intensities from the colonies arrayed on solid agar plates and all of these intensities can be
quantified using appropriate software. Deletion mutants that cause decreased GFP:RFP
ratios are indicative of genes that are required for activation of the promoter of interest while
deletion mutants causing increased GFP:RFP ratios reveal genes that repress the promoter of
interest.
117
In Chapter 2, I described the complete methodology for carrying out this type of
screen including particular details we discovered to generate high quality data. I describe the
SGA-selection media required and timeline for generating the output array as well as general
guidelines to consider when screening. I show step-by-step how we carry out the screen
using a CLN2 promoter-GFP reporter gene as an example, revealing that this screen can
distinguish the known transcriptional activator, Swi4, as the principal activator of CLN2
expression. Further, I show that normalization of GFP to the control RFP signal is
imperative to reliably detect Swi4 as the top activator of CLN2 expression compared to
considering the GFP signal alone. This information shows the utility of the approach and
will guide other researchers in adapting this type of screening to applications of their choice.
In Chapter 3, I describe the application of the above screening approach to probe the
yeast genome for new regulators of histone gene transcription, an important group of genes
whose expression is tightly regulated during S-phase of the cell cycle (Hereford et al., 1981).
In S. cerevisiae, each of the four histones are encoded by two different genes and histone
pairs are divergently transcribed from the same intergenic region so that the HTA1-HTB1 and
HTA2-HTB2 gene pairs encode histone H2A and H2B expression while HHT1-HHF1 and
HHT2-HHF2 encode H3 and H4 expression respectively. The promoter regions of three of
the four gene pairs (excluding HTA2-HTB2) contain a cis-acting negative regulatory DNA
sequence called NEG which is required for cell cycle regulation of histone transcripts and
repression through this site is mediated by the Asf1/HIR protein complex (Green et al., 2005;
Osley et al., 1986; Osley and Lycan, 1987; Prochasson et al., 2005; Xu et al., 1992). To
probe for new repressors and better characterize Asf1/HIR mediated repression of these
genes, as well as identify potential activators, I carried out a R-SGA screen with the HTA1
118
promoter fused to GFP. I discovered that the histone H3-H4 chaperone protein Rtt106 is a
novel repressor of HTA1 transcription while other proteins like Rtt109 and Yta7 play an
activating role.
To characterize the roles these proteins play, I worked with my collaborators to
perform a series of experiments including chromatin immunoprecipitation, transcript
profiling, genome-wide nucleosome positioning and co-purification. These experiments
revealed that Rtt106 is a novel member of the Asf1/HIR pathway that regulates histone
transcription through the NEG site. Like HIR and Asf1, Rtt106 is required for proper cell
cycle regulation of histone transcripts. Our genome-wide nucleosome positioning analysis
revealed that the absence of Rtt106 or Hir1 renders histone promoters containing the NEG
sequence nucleosome free, indicating that cell cycle regulation of histone transcription is
controlled by proper nucleosome occupancy at the promoter. Because Rtt106 is
constitutively localized at the promoter throughout the cell cycle, we propose an activating
factor is required to overcome Asf1/HIR/Rtt106 mediated repression of histone genes. One
gene that is required for activation of histone genes encodes the histone H3 K56 specific
HAT, Rtt109, which seems to play some role in countering Rtt106 and Hir1 mediated
repression. We also show that the bromodomain containing protein Yta7 acts as a boundary
element at histone gene promoters by restricting the localization of Rtt106 specifically to the
NEG-containing region of the promoter and excluding it from the coding regions. These
findings illustrate the power of an unbiased, functional genomics approach for identifying
new transcriptional regulators. We identified and characterized novel members of the
Asf1/HIR pathway and our genome-wide nucleosome positioning experiments suggest that
119
the Rtt106/Asf1/HIR/Yta7 pathway we characterized is acting at promoters broadly in the
genome.
4.2 Future directions
4.2.1 Characterizing the Rtt106/Asf1/HIR/Yta7 pathway genome-wide
Our efforts so far have uncovered a novel pathway where the Rtt106/Asf1/HIR
proteins repress transcription at histone promoters and Rtt106 localization is restricted to the
promoter regions and excluded from the coding regions by the presence of Yta7. Because
we carried out nucleosome positioning experiments genome-wide in the absence of Rtt106
and Hir1, we were able to identify hundreds of promoters that appear to be nucleosome-free
when these genes are deleted, suggesting that the Rtt106/Asf1/HIR/Yta7 pathway defines a
previously unappreciated transcriptional regulatory mechanism that is broadly applied across
the genome. To test this idea, I propose ChIP-seq experiments to determine the localization
patterns of these proteins on DNA. Until recently, this type of experiment has been carried
out by hybridizing ChIP-enriched DNA to a microarray to identify regions occupied by a
protein of interest. However, the advent of next generation sequencing technologies has
made ChIP-seq experiments feasible. Sequencing allows greater resolution for mapping
protein-DNA interactions and for small genomes like yeast, multiplexing experiments in
single sequencing runs allows cost-effective generation of large data sets (Lefrancois et al.,
2009).
We have shown efficient cross-linking of TAP-tagged versions of Rtt106, Hir1 and
Yta7 to the regulatory regions of histone genes and these strains can be used for the ChIP-seq
experiments. If Rtt106, HIR and Yta7 are collaborating to create regions of repressed
120
chromatin in promoters of genes, which is suggested from our previous nucleosome
positioning data, I expect to find these proteins localizing to the same promoters across the
genome. Our previous work also revealed that Yta7 is a boundary protein at the cell cycle-
regulated histone genes. When YTA7 is deleted, Rtt106 spreads laterally from the NEG-site
in the promoter and into the ORF. Thus, to determine whether Yta7 is acting as a boundary
element at other active promoters, ChIP-seq on the Rtt106-TAP yta7Δ strain could be
performed. I expect that many promoters will be occupied by Rtt106 and Hir1 and that these
promoters will also be targeted by Yta7. In the absence of Yta7, I expect to see spreading of
Rtt106, similar to what we observed at the HTA1 locus.
Interestingly, localization of Rtt106 and Yta7 to the NEG site of histone promoters is
dependent on HIR and this NEG sequence is only present in histone gene promoters. It is
possible that at other promoters, Rtt106 is recruited by other factors independently of HIR in
which case ChIP-seq experiments would reveal HIR-dependent and HIR-independent
pathways of Rtt106 transcriptional regulation at promoters. This might particularly be true
for promoters that are independent of cell cycle regulation, since the NEG site is required to
confer proper S-phase specific expression of histone genes (Osley, 1991; Osley et al., 1986).
To determine experimentally if this is the case, I will carry out ChIP-seq experiments on the
Rtt106-TAP hir1Δ strain. The results of this experiment will reveal whether Rtt106 can
localize to promoters in the genome in the absence of HIR1. Because a single sequencing
run generates many more sequence reads than is required, four different ChIP-enriched DNA
samples can be multiplexed and sequenced in a single reaction using an Illumina Genome
Analyzer II sequencing platform (Lefrancois et al., 2009). To further characterize HIR-
independent Rtt106-regulated promoters, I propose to fuse those promoters that I discover in
121
my ChIP-seq experiments to GFP and carry out an R-SGA screen to identify other regulators
of those promoters that might be responsible for recruiting Rtt106.
Since our results show a role for Rtt106 in regulating transcription, gene expression
microarray experiments will be carried out in rtt106Δ, hir1Δ and yta7Δ mutants to correlate
promoter localization of these proteins with transcriptional output. These data will also be
compared to our nucleosome positioning data already generated. Thus, I plan to characterize
the Rtt106-Hir-Yta7 pathway by overlapping data from ChIP-seq, gene expression
microarrays and nucleosome occupancy experiments.
4.2.2 Characterizing protein domains in Rtt106 required for function with HIR
Our analysis of Rtt106 revealed a role for this protein in repressing transcription of
the cell cycle-histone genes. As noted in Chapter 3, Rtt106 has higher affinity for K56Ac
histone H3 (Li et al., 2008), a chromatin mark that is abundant in S-phase of the cell cycle by
action of the HAT Rtt109 (Collins et al., 2007; Driscoll et al., 2007; Han et al., 2007a;
Ozdemir et al., 2005; Xu et al., 2005). A PH-like domain was discovered in Rtt106 from
residues 195 to 301 that mediates binding to H3 K56Ac (Li et al., 2008). Because K56Ac is
known to be important for activation of histone gene transcription (Xu et al., 2005), it is of
interest to understand if the PH-like domain in Rtt106 plays a role in regulation of histone
genes. It is possible that Rtt109 acetylates H3 K56 and this mark is incorporated into the
histone promoter region by Rtt106 to allow activation of transcription transiently during S-
phase which is rapidly repressed by incorporation of non-acetylated H3 K56 histones during
other phases of the cell cycle by Rtt106.
122
To test this hypothesis, I propose to carry out experiments with a strain where the PH-
like domain of Rtt106 is deleted (RTT106Δ196-301) (Li et al., 2008). First, I will test
endogenous HTA1 transcript levels in the RTT106Δ196-301 strain using qPCR (see Chapter
3) to determine if there remains derepression of HTA1 transcripts similar to the rtt106Δ
strain. If the above hypothesis is correct, transcript levels may be reduced in the rtt106Δ196-
301 strain (possibly similar to the rtt109Δ strain) compared to the rtt106Δ strain because of a
failure to incorporate K56Ac histone H3 into the promoter. Next, I will test H3 K56Ac
levels specifically at the HTA1 promoter using ChIP with an antibody that recognizes H3
K56Ac in both the rtt106Δ and rtt106Δ196-301 strains and compare these levels to histone
H3 levels using an antibody specific for histone H3. In both strains, K56Ac maybe be
reduced because of the absence of the PH domain of Rtt106 at the promoter but H3 levels
may not be completely reduced in the rtt106Δ196-301 relative to rtt106Δ, which causes the
promoter to become nucleosome free. In the rtt106Δ196-301 strain, there may be other
protein domains that still allow proper nucleosome assembly of unacetylated H3 K56
histones to still allow repression outside of S-phase so that the promoter does not become
nucleosome-free and constitutively active like in the rtt106Δ strain. An important control is
to test whether a TAP-tagged version of the Rtt106-PH domain mutant protein still localizes
to the NEG-containing region of the histone promoter using the ChIP assay described in
Chapter 3. Furthermore, it will be important to test whether its localization is still dependent
on HIR, as for Rtt106-TAP.
123
4.2.3 Screening overexpression arrays
The screening approach I described here makes use of a collection of yeast deletion
mutants. However, because ~1000 genes are essential (Giaever et al., 2002), the effect that
mutation of these genes causes on reporter genes cannot be scored in the present screening
format. One way to identify regulatory pathways controlled by essential genes is to examine
the consequence of gene overexpression on a reporter gene of interest. In this case, an array
of yeast strains where each yeast colony represents overexpression of a yeast protein is useful
(Sopko et al., 2006). Each colony on this array harbours a high-copy plasmid where the
GAL1/10 promoter, which is induced in the presence of galactose, drives expression of a
different ORF (Sopko et al., 2006). This array can be manipulated using the SGA
methodology to combine GFP and RFP reporter genes with overexpression of each ORF
when grown in the presence of galactose. Opposite effects to gene-deletions are likely to
result such that increased GFP:RFP would indicate the overexpressed gene is an activator of
the promoter while decreased GFP:RFP would indicate the overexpressed gene is a repressor
of the promoter (Figure 4-1).
124
Figure 4-1: Overexpression screen to identify regulators of a promoter of interest. The SGA methodology is used to combine a promoter-GFP and a control promoter-RFP gene into an ordered array of galactose inducible overexpression strains. The output array is replicated on media containing galactose to induce overexpression of each ORF and colonies arrayed on agar plates are assayed using the typhoon fluorescence scanner. Since overexpression of most ORFs will not affect gene expression, the combination of GFP and RFP in those strains will result in a yellow colour. Increased GFP compared to RFP will indicate the overexpressed gene is an activator of the promoter of interest while decreased GFP compared to RFP will indicate the overexpressed gene is a repressor of the promoter of interest.
125
Previous work has shown that deletion of many transcription factors results in little
effect on target genes examined by gene expression microarrays, likely because the
transcription factor is inactive under the conditions tested (Chua et al., 2006). However,
overexpression in many cases resulted in a gene expression pattern above noise indicating
that artificially activating expression of these proteins bypasses the need for specific
activating conditions and can be used to identify target genes of transcription factors (Chua et
al., 2006). Similarly, by overexpressing genes it would be possible to identify activating
pathways of a promoter that may not normally be active in the conditions tested. Thus,
overexpression screens are advantageous because they allow analysis of essential genes,
regulatory proteins that are only expressed under specific conditions can be assayed for an
activating role on promoters of interest and they can provide complementary data to deletion
screens.
4.2.4 Increasing throughput of reporter-gene analysis using pooled screens in yeast and
higher eukaryotes
I developed a novel approach to study gene regulation by combining fluorescent
reporter genes, the SGA approach and a simple assay for detecting fluorescence from yeast
colonies arrayed on agar plates. Although this approach has proven useful for studying gene
regulatory pathways, the plate-based colony assay is generally not adaptable to higher
eukaryotes. I propose to develop and validate a methodology to identify trans-acting factors
of promoter-reporter constructs using pooled cultures that could be adapted to any organism
for which an appropriate gene disruption library exists. Because of the potential for
126
exploring the possibility of new methodologies in S. cerevisiae, I propose to first test this
approach in yeast.
As I described in Chapter 1, the yeast deletion library contains each yeast ORF
knocked out with a kanamycin resistance cassette. In addition, each knockout cassette is
flanked by unique sequences that are used as strain identifiers or molecular barcodes and all
mutants contain a universal sequence that allows a single primer set to amplify each unique
barcode (Giaever et al., 2002 and Figure 1-6). If the entire collection is pooled into a single
culture and treated with a particular condition (for example drug treatment), deletion strains
that are sensitive to treatment are under-represented in the population. DNA is prepared
from the pooled culture, PCR-amplified with the universal primer set and hybridized to a
microarray that contains oligonucleotide probes homologous to each molecular barcode.
Positions on the microarray that no longer show signal after hybridization are indicative of
mutants sensitive to the particular treatment. This type of strategy has been used to
quantitatively monitor the deletion collection for strains that show a fitness defect when
grown in rich-media or under various conditions (Giaever et al., 2002).
Since these barcodes allows identification of deletion mutants in a mixed population,
the potential for combining this pooled strategy with fluorescent reporter genes and
fluorescence activated cell sorting (FACS) exists. In this case, a promoter-GFP reporter gene
can be introduced into the collection of deletion mutants using the SGA approach or by
directly transforming the reporter plasmid into a pooled culture of all viable deletion strains.
The population of cells is pooled and grown in appropriate selection media and subject to
FACS so that the brightest cells in the population are sorted into one sub-population and the
dimmest cells are sorted into a different sub-population (Figure 4-2). Flow cytometry has
127
been used to make precise measurements of GFP-tagged proteins from single cells, which
indicates the feasibility of the approach described here for detecting fine cell-to-cell changes
in GFP expression (Newman et al., 2006). The brightest cells expressing high levels of GFP
likely contain deletion mutants that are required to repress the promoter of interest while the
dimmest cells likely contain activators of the promoter. To identify which mutants are in
each population, DNA prepared from each pool is hybridized to a barcode microarray and
compared to hybridization signal from an unsorted population (Figure 4-2). This type of
approach could be used to determine how the promoter-reporter gene responds to different
environmental conditions or drug treatments (particularly when drugs are scarce) in each
mutant background and should aid in discovery of new pathways of gene regulation.
128
Figure 4-2: Reporter screening using barcoded gene disruption libraries and FACS. A pooled culture of cells with a promoter-GFP reporter gene combined with each yeast deletion mutant is subject to FACS to physically sort cells based on the intensity of GFP signal. Mutants present in the brightest and dimmest sub-populations after sorting are identified by hybridization to a barcode microarray and normalized to the hybridization signal from the unsorted population, which defines the mutants present in the initial pooled culture.
129
In human cells and other eukaryotes, barcoded RNAi libraries exist to perturb
expression of each gene (Moffat et al., 2006). Reporter genes of interest could be stably
expressed in cell lines of choice and combined with the RNAi library. Similar to the above
experiment, FACS cell sorting could be used to physically sort bright and dim cells from the
population and RNAi molecular barcodes detected to determine which gene is targeted.
Pooled RNAi barcode screens have been carried out previously on a genome-scale and
identified a number of genes required for cell proliferation in different tumour types
(Schlabach et al., 2008; Silva et al., 2008). These types of screens should prove useful for
studying transcriptional regulatory pathways that are perturbed in cancer cell types.
4.3 Overall significance
The advent of DNA microarrays has revolutionized the gene expression field and has
led to major discoveries on regulator-gene interactions. From parallel transcript profiling,
genome-wide localization of proteins and histone modifications on chromatin, maps of
nucleosome positions and DNA sequences bound by regulatory proteins, array-based
approaches have allowed detailed analysis of mechanisms that control how the genome is
regulated. Now that next generation sequencing applications are becoming commonplace,
gene expression regulatory pathways are being examined with unprecedented resolution.
With the work presented in this Thesis, it is now possible to carry out systematic and
quantitative gene expression reporter screens using array-based reagents in yeast, a
methodology that was previously lacking. Based on this work, I propose a new approach
that is based on sorting cells with fluorescent reporter genes that should be adaptable to
130
higher organisms and aid in the discovery of new transcription factor pathways that have
previously been unappreciated.
It is clear that even though an enormous amount of gene expression data has been
generated, there remain many more regulatory pathways to be discovered, even in well
studied organisms like S. cerevisiae. My work has shown the power of systematic reporter
screens for linking protein function to transcriptional regulation that would otherwise remain
mysterious. However, for complete elucidation and characterization of transcription factor
pathways, no one approach should be considered superior. Instead a combined approach
utilizing various functional genomic techniques and tools should be carried out to fully
realize the goal of defining all pathways that control gene expression.
131
References Adkins, M. W., Howar, S. R., and Tyler, J. K. (2004). Chromatin disassembly mediated by
the histone chaperone Asf1 is essential for transcriptional activation of the yeast
PHO5 and PHO8 genes. Mol Cell 14, 657-666.
Adkins, M. W., Williams, S. K., Linger, J., and Tyler, J. K. (2007). Chromatin disassembly
from the PHO5 promoter is essential for the recruitment of the general transcription
machinery and coactivators. Mol Cell Biol 27, 6372-6382.
Andrews, B. J., and Herskowitz, I. (1989). Identification of a DNA binding factor involved
in cell-cycle control of the yeast HO gene. Cell 57, 21-29.
Andrews, B. J., and Moore, L. A. (1992). Interaction of the yeast Swi4 and Swi6 cell cycle
regulatory proteins in vitro. Proc Natl Acad Sci U S A 89, 11852-11856.
Baba, T., Ara, T., Hasegawa, M., Takai, Y., Okumura, Y., Baba, M., Datsenko, K. A.,
Tomita, M., Wanner, B. L., and Mori, H. (2006). Construction of Escherichia coli K-
12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2,
2006 0008.
Bader, G. D., Heilbut, A., Andrews, B., Tyers, M., Hughes, T., and Boone, C. (2003).
Functional genomics and proteomics: charting a multidimensional map of the yeast
cell. Trends Cell Biol 13, 344-356.
Badis, G., Chan, E. T., van Bakel, H., Pena-Castillo, L., Tillo, D., Tsui, K., Carlson, C. D.,
Gossett, A. J., Hasinoff, M. J., Warren, C. L., et al. (2008). A library of yeast
transcription factor motifs reveals a widespread function for Rsc3 in targeting
nucleosome exclusion at promoters. Mol Cell 32, 878-887.
Bammler, T., Beyer, R. P., Bhattacharya, S., Boorman, G. A., Boyles, A., Bradford, B. U.,
132
Bumgarner, R. E., Bushel, P. R., Chaturvedi, K., Choi, D., et al. (2005).
Standardizing global gene expression analysis between laboratories and across
platforms. Nat Methods 2, 351-356.
Barbaric, S., Munsterkotter, M., Svaren, J., and Horz, W. (1996). The homeodomain protein
Pho2 and the basic-helix-loop-helix protein Pho4 bind DNA cooperatively at the
yeast PHO5 promoter. Nucleic Acids Res 24, 4479-4486.
Berger, M. F., and Bulyk, M. L. (2006). Protein binding microarrays (PBMs) for rapid, high-
throughput characterization of the sequence specificities of DNA binding proteins.
Methods Mol Biol 338, 245-260.
Berger, M. F., Philippakis, A. A., Qureshi, A. M., He, F. S., Estep, P. W., 3rd, and Bulyk, M.
L. (2006). Compact, universal DNA microarrays to comprehensively determine
transcription-factor binding site specificities. Nat Biotechnol 24, 1429-1435.
Bernstein, B. E., Liu, C. L., Humphrey, E. L., Perlstein, E. O., and Schreiber, S. L. (2004).
Global nucleosome occupancy in yeast. Genome Biol 5, R62.
Boeger, H., Griesenbeck, J., Strattan, J. S., and Kornberg, R. D. (2003). Nucleosomes unfold
completely at a transcriptionally active promoter. Mol Cell 11, 1587-1598.
Borneman, A. R., Gianoulis, T. A., Zhang, Z. D., Yu, H., Rozowsky, J., Seringhaus, M. R.,
Wang, L. Y., Gerstein, M., and Snyder, M. (2007). Divergence of transcription factor
binding sites across related yeast species. Science 317, 815-819.
Boutros, M., Kiger, A. A., Armknecht, S., Kerr, K., Hild, M., Koch, B., Haas, S. A., Paro, R.,
and Perrimon, N. (2004). Genome-wide RNAi analysis of growth and viability in
Drosophila cells. Science 303, 832-835.
Breeden, L., and Nasmyth, K. (1987). Cell cycle control of the yeast HO gene: cis- and trans-
133
acting regulators. Cell 48, 389-397.
Brutlag, D., Schlehuber, C., and Bonner, J. (1969). Properties of formaldehyde-treated
nucleohistone. Biochemistry 8, 3214-3218.
Butland, G., Babu, M., Diaz-Mejia, J. J., Bohdana, F., Phanse, S., Gold, B., Yang, W., Li, J.,
Gagarinova, A. G., Pogoutse, O., et al. (2008). eSGA: E. coli synthetic genetic array
analysis. Nat Methods 5, 789-795.
Byrne, A. B., Weirauch, M. T., Wong, V., Koeva, M., Dixon, S. J., Stuart, J. M., and Roy, P.
J. (2007). A global analysis of genetic interactions in Caenorhabditis elegans. J Biol
6, 8.
Celic, I., Masumoto, H., Griffith, W. P., Meluh, P., Cotter, R. J., Boeke, J. D., and Verreault,
A. (2006). The sirtuins hst3 and Hst4p preserve genome integrity by controlling
histone h3 lysine 56 deacetylation. Curr Biol 16, 1280-1289.
Cho, R. J., Campbell, M. J., Winzeler, E. A., Steinmetz, L., Conway, A., Wodicka, L.,
Wolfsberg, T. G., Gabrielian, A. E., Landsman, D., Lockhart, D. J., and Davis, R. W.
(1998). A genome-wide transcriptional analysis of the mitotic cell cycle. Mol Cell 2,
65-73.
Cho, R. J., Huang, M., Campbell, M. J., Dong, H., Steinmetz, L., Sapinoso, L., Hampton, G.,
Elledge, S. J., Davis, R. W., and Lockhart, D. J. (2001). Transcriptional regulation
and function during the human cell cycle. Nat Genet 27, 48-54.
Chua, G., Morris, Q. D., Sopko, R., Robinson, M. D., Ryan, O., Chan, E. T., Frey, B. J.,
Andrews, B. J., Boone, C., and Hughes, T. R. (2006). Identifying transcription factor
functions and targets by phenotypic activation. Proc Natl Acad Sci U S A 103,
12045-12050.
134
Cleveland, W. S. (1979). Robust locally weighted regression and smoothing scatterplots.
Journal of the American Statistical Association 74, 829-836.
Collins, S. R., Miller, K. M., Maas, N. L., Roguev, A., Fillingham, J., Chu, C. S., Schuldiner,
M., Gebbia, M., Recht, J., Shales, M., et al. (2007). Functional dissection of protein
complexes involved in yeast chromosome biology using a genetic interaction map.
Nature 446, 806-810.
Cosma, M. P., Tanaka, T., and Nasmyth, K. (1999). Ordered recruitment of transcription and
chromatin remodeling factors to a cell cycle- and developmentally regulated
promoter. Cell 97, 299-311.
Costanzo, M., Nishikawa, J. L., Tang, X., Millman, J. S., Schub, O., Breitkreuz, K., Dewar,
D., Rupes, I., Andrews, B., and Tyers, M. (2004). CDK activity antagonizes Whi5, an
inhibitor of G1/S transcription in yeast. Cell 117, 899-913.
David, L., Huber, W., Granovskaia, M., Toedling, J., Palm, C. J., Bofkin, L., Jones, T.,
Davis, R. W., and Steinmetz, L. M. (2006). A high-resolution map of transcription in
the yeast genome. Proc Natl Acad Sci U S A 103, 5320-5325.
Davierwala, A. P., Haynes, J., Li, Z., Brost, R. L., Robinson, M. D., Yu, L., Mnaimneh, S.,
Ding, H., Zhu, H., Chen, Y., et al. (2005). The synthetic genetic interaction spectrum
of essential genes. Nat Genet 37, 1147-1152.
De Koning, L., Corpet, A., Haber, J. E., and Almouzni, G. (2007). Histone chaperones: an
escort network regulating histone traffic. Nat Struct Mol Biol 14, 997-1007.
de Reynies, A., Geromin, D., Cayuela, J. M., Petel, F., Dessen, P., Sigaux, F., and
Rickman, D. S. (2006). Comparison of the latest commercial short and long
oligonucleotide microarray technologies. BMC Genomics 7, 51.
135
De Virgilio, C., Burckert, N., Bell, W., Jeno, P., Boller, T., and Wiemken, A. (1993).
Disruption of TPS2, the gene encoding the 100-kDa subunit of the trehalose-6-
phosphate synthase/phosphatase complex in Saccharomyces cerevisiae, causes
accumulation of trehalose-6-phosphate and loss of trehalose-6-phosphate phosphatase
activity. Eur J Biochem 212, 315-323.
Dimova, D., Nackerdien, Z., Furgeson, S., Eguchi, S., and Osley, M. A. (1999). A role for
transcriptional repressors in targeting the yeast Swi/Snf complex. Mol Cell 4, 75-83.
Dixon, S. J., Fedyshyn, Y., Koh, J. L., Prasad, T. S., Chahwan, C., Chua, G., Toufighi, K.,
Baryshnikova, A., Hayles, J., Hoe, K. L., et al. (2008). Significant conservation of
synthetic lethal genetic interaction networks between distantly related eukaryotes.
Proc Natl Acad Sci U S A 105, 16653-16658.
Dollard, C., Ricupero-Hovasse, S. L., Natsoulis, G., Boeke, J. D., and Winston, F. (1994).
SPT10 and SPT21 are required for transcription of particular histone genes in
Saccharomyces cerevisiae. Mol Cell Biol 14, 5223-5228.
Driscoll, R., Hudson, A., and Jackson, S. P. (2007). Yeast Rtt109 promotes genome stability
by acetylating histone H3 on lysine 56. Science 315, 649-652.
Eriksson, P. R., Mendiratta, G., McLaughlin, N. B., Wolfsberg, T. G., Marino- Ramirez,
L., Pompa, T. A., Jainerin, M., Landsman, D., Shen, C. H., and Clark, D. J.
(2005). Global regulation by the yeast Spt10 protein is mediated through chromatin
structure and the histone upstream activating sequence elements. Mol Cell Biol 25,
9127-9137.
Evans, T., Rosenthal, E. T., Youngblom, J., Distel, D., and Hunt, T. (1983). Cyclin: a protein
specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage
136
division. Cell 33, 389-396.
Fillingham, J., Recht, J., Silva, A. C., Suter, B., Emili, A., Stagljar, I., Krogan, N. J., Allis, C.
D., Keogh, M. C., and Greenblatt, J. F. (2008). Chaperone control of the activity and
specificity of the histone H3 acetyltransferase Rtt109. Mol Cell Biol 28, 4342-4353.
Giaever, G., Chu, A. M., Ni, L., Connelly, C., Riles, L., Veronneau, S., Dow, S., Lucau-
Danila, A., Anderson, K., Andre, B., et al. (2002). Functional profiling of the
Saccharomyces cerevisiae genome. Nature 418, 387-391.
Giresi, P. G., and Lieb, J. D. (2009). Isolation of active regulatory elements from eukaryotic
chromatin using FAIRE (Formaldehyde Assisted Isolation of Regulatory Elements).
Methods.
Goldstein, A. L., and McCusker, J. H. (1999). Three new dominant drug resistance cassettes
for gene disruption in Saccharomyces cerevisiae. Yeast 15, 1541-1553.
Gradolatto, A., Rogers, R. S., Lavender, H., Taverna, S. D., Allis, C. D., Aitchison, J. D., and
Tackett, A. J. (2008). Saccharomyces cerevisiae Yta7 regulates histone gene
expression. Genetics 179, 291-304.
Green, E. M., Antczak, A. J., Bailey, A. O., Franco, A. A., Wu, K. J., Yates, J. R., 3rd, and
Kaufman, P. D. (2005). Replication-independent histone deposition by the HIR
complex and Asf1. Curr Biol 15, 2044-2049.
Guarente, L., Yocum, R. R., and Gifford, P. (1982). A GAL10-CYC1 hybrid yeast promoter
identifies the GAL4 regulatory region as an upstream site. Proc Natl Acad Sci U S A
79, 7410-7414.
Gunjan, A., Paik, J., and Verreault, A. (2005). Regulation of histone synthesis and
nucleosome assembly. Biochimie 87, 625-635.
137
Gunjan, A., and Verreault, A. (2003). A Rad53 kinase-dependent surveillance mechanism
that regulates histone protein levels in S. cerevisiae. Cell 115, 537-549.
Hadwiger, J. A., Wittenberg, C., Richardson, H. E., de Barros Lopes, M., Reed, S. I. (1989).
A family of cyclin homologs that control the G1 phase in yeast. Proc Natl Acad Sci U
S A 86, 6255-6259.
Han, J., Zhou, H., Horazdovsky, B., Zhang, K., Xu, R. M., and Zhang, Z. (2007a). Rtt109
acetylates histone H3 lysine 56 and functions in DNA replication. Science 315, 653-
655.
Han, J., Zhou, H., Li, Z., Xu, R. M., and Zhang, Z. (2007b). The Rtt109-Vps75 histone
acetyltransferase complex acetylates non-nucleosomal histone H3. J Biol Chem 282,
14158-14164.
Harbison, C. T., Gordon, D. B., Lee, T. I., Rinaldi, N. J., Macisaac, K. D., Danford, T. W.,
Hannett, N. M., Tagne, J. B., Reynolds, D. B., Yoo, J., et al. (2004). Transcriptional
regulatory code of a eukaryotic genome. Nature 431, 99-104.
Heintzman, N. D., Stuart, R. K., Hon, G., Fu, Y., Ching, C. W., Hawkins, R. D., Barrera, L.
O., Van Calcar, S., Qu, C., Ching, K. A., et al. (2007). Distinct and predictive
chromatin signatures of transcriptional promoters and enhancers in the human
genome. Nat Genet 39, 311-318.
Hereford, L. M., Osley, M. A., Ludwig, T. R., 2nd, and McLaughlin, C. S. (1981). Cell-cycle
regulation of yeast histone mRNA. Cell 24, 367-375.
Hess, D., Liu, B., Roan, N. R., Sternglanz, R., and Winston, F. (2004). Spt10-dependent
transcriptional activation in Saccharomyces cerevisiae requires both the Spt10
acetyltransferase domain and Spt21. Mol Cell Biol 24, 135-143.
138
Hess, D., and Winston, F. (2005). Evidence that Spt10 and Spt21 of Saccharomyces
cerevisiae play distinct roles in vivo and functionally interact with MCB-binding
factor, SCB-binding factor and Snf1. Genetics 170, 87-94.
Hiraga, S., Botsios, S., and Donaldson, A. D. (2008). Histone H3 lysine 56 acetylation by
Rtt109 is crucial for chromosome positioning. J Cell Biol 183, 641-651.
Horak, C. E., Luscombe, N. M., Qian, J., Bertone, P., Piccirrillo, S., Gerstein, M., and
Snyder, M. (2002). Complex transcriptional circuitry at the G1/S transition in
Saccharomyces cerevisiae. Genes Dev 16, 3017-3033.
Huang, S., Zhou, H., Katzmann, D., Hochstrasser, M., Atanasova, E., and Zhang, Z. (2005).
Rtt106p is a histone chaperone involved in heterochromatin-mediated silencing. Proc
Natl Acad Sci U S A 102, 13410-13415.
Huang, S., Zhou, H., Tarara, J., and Zhang, Z. (2007). A novel role for histone chaperones
CAF-1 and Rtt106p in heterochromatin silencing. Embo J 26, 2274-2283.
Hughes, J. D., Estep, P. W., Tavazoie, S., and Church, G. M. (2000a). Computational
identification of cis-regulatory elements associated with groups of functionally
related genes in Saccharomyces cerevisiae. J Mol Biol 296, 1205-1214.
Hughes, T. R., Mao, M., Jones, A. R., Burchard, J., Marton, M. J., Shannon, K. W.,
Lefkowitz, S. M., Ziman, M., Schelter, J. M., Meyer, M. R., et al. (2001). Expression
profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat
Biotechnol 19, 342-347.
Hughes, T. R., Marton, M. J., Jones, A. R., Roberts, C. J., Stoughton, R., Armour, C. D.,
Bennett, H. A., Coffey, E., Dai, H., He, Y. D., et al. (2000b). Functional discovery
via a compendium of expression profiles. Cell 102, 109-126.
139
Hughes, T. R., and Shoemaker, D. D. (2001). DNA microarrays for expression profiling.
Curr Opin Chem Biol 5, 21-25.
Huh, W. K., Falvo, J. V., Gerke, L. C., Carroll, A. S., Howson, R. W., Weissman, J. S., and
O'Shea, E. K. (2003). Global analysis of protein localization in budding yeast. Nature
425, 686-691.
Idnurm, A., Walton, F. J., Floyd, A., Reedy, J. L., and Heitman, J. (2009). Identification of
ENA1 as a virulence gene of the human pathogenic fungus Cryptococcus neoformans
through signature-tagged insertional mutagenesis. Eukaryot Cell 8, 315-326.
Imbeault, D., Gamar, L., Rufiange, A., Paquet, E., and Nourani, A. (2008). The Rtt106
histone chaperone is functionally linked to transcription elongation and is involved in
the regulation of spurious transcription from cryptic promoters in yeast. J Biol Chem
283, 27350-27354.
Ivessa, A. S., Lenzmeier, B. A., Bessler, J. B., Goudsouzian, L. K., Schnakenberg, S. L., and
Zakian, V. A. (2003). The Saccharomyces cerevisiae helicase Rrm3p facilitates
replication past nonhistone protein-DNA complexes. Mol Cell 12, 1525-1536.
Iyer, V. R., Horak, C. E., Scafe, C. S., Botstein, D., Snyder, M., and Brown, P. O. (2001).
Genomic binding sites of the yeast cell-cycle transcription factors SBF and MBF.
Nature 409, 533-538.
Jambunathan, N., Martinez, A. W., Robert, E. C., Agochukwu, N. B., Ibos, M. E., Dugas, S.
L., and Donze, D. (2005). Multiple bromodomain genes are involved in restricting the
spread of heterochromatic silencing at the Saccharomyces cerevisiae HMR-tRNA
boundary. Genetics 171, 913-922.
140
Jansen, R. P., Dowzer, C., Michaelis, C., Galova, M., and Nasmyth, K. (1996). Mother cell-
specific HO expression in budding yeast depends on the unconventional myosin
myo4p and other cytoplasmic proteins. Cell 84, 687-697.
Jensen, R., Sprague, G. F., Jr., and Herskowitz, I. (1983). Regulation of yeast mating-type
interconversion: feedback control of HO gene expression by the mating-type locus.
Proc Natl Acad Sci U S A 80, 3035-3039.
Kainth, P., Sassi, H. E., Pena-Castillo, L., Chua, G., Hughes, T. R., and Andrews, B. (2009).
Comprehensive genetic analysis of transcription factor pathways using a dual reporter
gene system in budding yeast. Methods.
Kamath, R. S., Fraser, A. G., Dong, Y., Poulin, G., Durbin, R., Gotta, M., Kanapin, A., Le
Bot, N., Moreno, S., Sohrmann, M., et al. (2003). Systematic functional analysis of
the Caenorhabditis elegans genome using RNAi. Nature 421, 231-237.
Kaplan, N., Moore, I. K., Fondufe-Mittendorf, Y., Gossett, A. J., Tillo, D., Field, Y.,
LeProust, E. M., Hughes, T. R., Lieb, J. D., Widom, J., and Segal, E. (2009). The
DNA-encoded nucleosome organization of a eukaryotic genome. Nature 458, 362-
366.
Kim, J., Bhinge, A. A., Morgan, X. C., and Iyer, V. R. (2005). Mapping DNA-protein
interactions in large genomes by sequence tag analysis of genomic enrichment. Nat
Methods 2, 47-53.
Kim, M., Ahn, S. H., Krogan, N. J., Greenblatt, J. F., and Buratowski, S. (2004). Transitions
in RNA polymerase II elongation complexes at the 3' ends of genes. Embo J 23, 354-
364.
141
Koch, C., Moll, T., Neuberg, M., Ahorn, H., and Nasmyth, K. (1993). A role for the
transcription factors Mbp1 and Swi4 in progression from G1 to S phase. Science 261,
1551-1557.
Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128, 693-705.
Krogan, N. J., Cagney, G., Yu, H., Zhong, G., Guo, X., Ignatchenko, A., Li, J., Pu, S., Datta,
N., Tikuisis, A. P., et al. (2006). Global landscape of protein complexes in the yeast
Saccharomyces cerevisiae. Nature 440, 637-643.
Lamb, J., Ramaswamy, S., Ford, H. L., Contreras, B., Martinez, R. V., Kittrell, F. S.,
Zahnow, C. A., Patterson, N., Golub, T. R., and Ewen, M. E. (2003). A mechanism of
cyclin D1 action encoded in the patterns of gene expression in human cancer. Cell
114, 323-334.
Lambert, J. P., Mitchell, L., Rudner, A., Baetz, K., and Figeys, D. (2009). A novel
proteomics approach for the discovery of chromatin-associated protein networks. Mol
Cell Proteomics 8, 870-882.
Lamour, V., Lecluse, Y., Desmaze, C., Spector, M., Bodescot, M., Aurias, A., Osley, M. A.,
and Lipinski, M. (1995). A human homolog of the S. cerevisiae HIR1 and HIR2
transcriptional repressors cloned from the DiGeorge syndrome critical region. Hum
Mol Genet 4, 791-799.
Lashkari, D. A., DeRisi, J. L., McCusker, J. H., Namath, A. F., Gentile, C., Hwang, S. Y.,
Brown, P. O., and Davis, R. W. (1997). Yeast microarrays for genome wide parallel
genetic and gene expression analysis. Proc Natl Acad Sci U S A 94, 13057-13062.
142
Lee, C. K., Shibata, Y., Rao, B., Strahl, B. D., and Lieb, J. D. (2004). Evidence for
nucleosome depletion at active regulatory regions genome-wide. Nat Genet 36, 900-
905.
Lee, T. I., Rinaldi, N. J., Robert, F., Odom, D. T., Bar-Joseph, Z., Gerber, G. K., Hannett, N.
M., Harbison, C. T., Thompson, C. M., Simon, I., et al. (2002). Transcriptional
regulatory networks in Saccharomyces cerevisiae. Science 298, 799-804.
Lee, W., Tillo, D., Bray, N., Morse, R. H., Davis, R. W., Hughes, T. R., and Nislow, C.
(2007). A high-resolution atlas of nucleosome occupancy in yeast. Nat Genet 39,
1235-1244.
Lefrancois, P., Euskirchen, G. M., Auerbach, R. K., Rozowsky, J., Gibson, T., Yellman, C.
M., Gerstein, M., and Snyder, M. (2009). Efficient yeast ChIP-Seq using multiplex
short-read DNA sequencing. BMC Genomics 10, 37.
Lehner, B., Crombie, C., Tischler, J., Fortunato, A., and Fraser, A. G. (2006). Systematic
mapping of genetic interactions in Caenorhabditis elegans identifies common
modifiers of diverse signaling pathways. Nat Genet 38, 896-903.
Li, Q., Zhou, H., Wurtele, H., Davies, B., Horazdovsky, B., Verreault, A., and Zhang, Z.
(2008). Acetylation of histone H3 lysine 56 regulates replication-coupled nucleosome
assembly. Cell 134, 244-255.
Lin, Y. Y., Qi, Y., Lu, J. Y., Pan, X., Yuan, D. S., Zhao, Y., Bader, J. S., and Boeke, J. D.
(2008). A comprehensive synthetic genetic interaction network governing yeast
histone acetylation and deacetylation. Genes Dev 22, 2062-2074.
Lipshutz, R. J., Fodor, S. P., Gingeras, T. R., and Lockhart, D. J. (1999). High density
synthetic oligonucleotide arrays. Nat Genet 21, 20-24.
143
Lycan, D. E., Osley, M. A., and Hereford, L. M. (1987). Role of transcriptional and
posttranscriptional regulation in expression of histone genes in Saccharomyces
cerevisiae. Mol Cell Biol 7, 614-621.
Maas, N. L., Miller, K. M., DeFazio, L. G., and Toczyski, D. P. (2006). Cell cycle and
checkpoint regulation of histone H3 K56 acetylation by Hst3 and Hst4. Mol Cell 23,
109-119.
Masumoto, H., Hawke, D., Kobayashi, R., and Verreault, A. (2005). A role for cell-cycle-
regulated histone H3 lysine 56 acetylation in the DNA damage response. Nature 436,
294-298.
Mendenhall, M. D., and Hodge, A. E. (1998). Regulation of Cdc28 cyclin-dependent protein
kinase activity during the cell cycle of the yeast Saccharomyces cerevisiae. Microbiol
Mol Biol Rev 62, 1191-1243.
Mnaimneh, S., Davierwala, A. P., Haynes, J., Moffat, J., Peng, W. T., Zhang, W., Yang, X.,
Pootoolal, J., Chua, G., Lopez, A., et al. (2004). Exploration of essential gene
functions via titratable promoter alleles. Cell 118, 31-44.
Moffat, J., Grueneberg, D. A., Yang, X., Kim, S. Y., Kloepfer, A. M., Hinkle, G., Piqani, B.,
Eisenhaure, T. M., Luo, B., Grenier, J. K., et al. (2006). A lentiviral RNAi library for
human and mouse genes applied to an arrayed viral high-content screen. Cell 124,
1283-1298.
Nagy, P. L., Cleary, M. L., Brown, P. O., and Lieb, J. D. (2003). Genomewide demarcation
of RNA polymerase II transcription units revealed by physical fractionation of
chromatin. Proc Natl Acad Sci U S A 100, 6364-6369.
Nasmyth, K. (1983). Molecular analysis of a cell lineage. Nature 302, 670-676.
144
Nasmyth, K., and Dirick, L. (1991). The role of SWI4 and SWI6 in the activity of G1 cyclins
in yeast. Cell 66, 995-1013.
Newman, J. R., Ghaemmaghami, S., Ihmels, J., Breslow, D. K., Noble, M., DeRisi, J. L., and
Weissman, J. S. (2006). Single-cell proteomic analysis of S. cerevisiae reveals the
architecture of biological noise. Nature 441, 840-846.
Nourani, A., Robert, F., and Winston, F. (2006). Evidence that Spt2/Sin1, an HMG-like
factor, plays roles in transcription elongation, chromatin structure, and genome
stability in Saccharomyces cerevisiae. Mol Cell Biol 26, 1496-1509.
Ogas, J., Andrews, B. J., and Herskowitz, I. (1991). Transcriptional activation of CLN1,
CLN2, and a putative new G1 cyclin (HCS26) by SWI4, a positive regulator of G1-
specific transcription. Cell 66, 1015-1026.
Oliva, A., Rosebrock, A., Ferrezuelo, F., Pyne, S., Chen, H., Skiena, S., Futcher, B., and
Leatherwood, J. (2005). The cell cycle-regulated genes of Schizosaccharomyces
pombe. PLoS Biol 3, e225.
Orlando, V. (2000). Mapping chromosomal proteins in vivo by formaldehyde-crosslinked-
chromatin immunoprecipitation. Trends Biochem Sci 25, 99-104.
Osley, M. A. (1991). The regulation of histone synthesis in the cell cycle. Annu Rev
Biochem 60, 827-861.
Osley, M. A., Gould, J., Kim, S., Kane, M. Y., and Hereford, L. (1986). Identification of
sequences in a yeast histone promoter involved in periodic transcription. Cell 45,
537-544.
Osley, M. A., and Lycan, D. (1987). Trans-acting regulatory mutations that alter
transcription of Saccharomyces cerevisiae histone genes. Mol Cell Biol 7, 4204-4210.
145
Ozdemir, A., Spicuglia, S., Lasonder, E., Vermeulen, M., Campsteijn, C., Stunnenberg, H.
G., and Logie, C. (2005). Characterization of lysine 56 of histone H3 as an
acetylation site in Saccharomyces cerevisiae. J Biol Chem 280, 25949-25952.
Pan, X., Yuan, D. S., Xiang, D., Wang, X., Sookhai-Mahadeo, S., Bader, J. S., Hieter, P.,
Spencer, F., and Boeke, J. D. (2004). A robust toolkit for functional profiling of the
yeast genome. Mol Cell 16, 487-496.
Peng, X., Karuturi, R. K., Miller, L. D., Lin, K., Jia, Y., Kondu, P., Wang, L., Wong, L. S.,
Liu, E. T., Balasubramanian, M. K., and Liu, J. (2005). Identification of cell cycle-
regulated genes in fission yeast. Mol Biol Cell 16, 1026-1042.
Perou, C. M., Jeffrey, S. S., van de Rijn, M., Rees, C. A., Eisen, M. B., Ross, D. T.,
Pergamenschikov, A., Williams, C. F., Zhu, S. X., Lee, J. C., et al. (1999).
Distinctive gene expression patterns in human mammary epithelial cells and breast
cancers. Proc Natl Acad Sci U S A 96, 9212-9217.
Pokholok, D. K., Harbison, C. T., Levine, S., Cole, M., Hannett, N. M., Lee, T. I., Bell, G.
W., Walker, K., Rolfe, P. A., Herbolsheimer, E., et al. (2005). Genome-wide map of
nucleosome acetylation and methylation in yeast. Cell 122, 517-527.
Polach, K. J., and Widom, J. (1995). Mechanism of protein access to specific DNA
sequences in chromatin: a dynamic equilibrium model for gene regulation. J Mol Biol
254, 130-149.
Pramila, T., Wu, W., Miles, S., Noble, W. S., and Breeden, L. L. (2006). The Forkhead
transcription factor Hcm1 regulates chromosome segregation genes and fills the S-
phase gap in the transcriptional circuitry of the cell cycle. Genes Dev 20, 2266-2278.
146
Prochasson, P., Florens, L., Swanson, S. K., Washburn, M. P., and Workman, J. L. (2005).
The HIR corepressor complex binds to nucleosomes generating a distinct
protein/DNA complex resistant to remodeling by SWI/SNF. Genes Dev 19, 2534-
2539.
Ramsay, G. (1998). DNA chips: state-of-the art. Nat Biotechnol 16, 40-44.
Recht, J., Tsubota, T., Tanny, J. C., Diaz, R. L., Berger, J. M., Zhang, X., Garcia, B. A.,
Shabanowitz, J., Burlingame, A. L., Hunt, D. F., et al. (2006). Histone chaperone
Asf1 is required for histone H3 lysine 56 acetylation, a modification associated with
S phase in mitosis and meiosis. Proc Natl Acad Sci U S A 103, 6988-6993.
Reinke, H., and Horz, W. (2003). Histones are first hyperacetylated and then lose contact
with the activated PHO5 promoter. Mol Cell 11, 1599-1607.
Ren, B., Robert, F., Wyrick, J. J., Aparicio, O., Jennings, E. G., Simon, I., Zeitlinger, J.,
Schreiber, J., Hannett, N., Kanin, E., et al. (2000). Genome-wide location and
function of DNA binding proteins. Science 290, 2306-2309.
Roguev, A., Wiren, M., Weissman, J. S., and Krogan, N. J. (2007). High-throughput genetic
interaction mapping in the fission yeast Schizosaccharomyces pombe. Nat Methods 4,
861-866.
Rufiange, A., Jacques, P. E., Bhat, W., Robert, F., and Nourani, A. (2007). Genome-wide
replication-independent histone H3 exchange occurs predominantly at promoters and
implicates H3 K56 acetylation and Asf1. Mol Cell 27, 393-405.
Rustici, G., Mata, J., Kivinen, K., Lio, P., Penkett, C. J., Burns, G., Hayles, J., Brazma, A.,
Nurse, P., and Bahler, J. (2004). Periodic gene expression program of the fission
yeast cell cycle. Nat Genet 36, 809-817.
147
Sassi, H. E., Bastajian, N., Kainth, P., and Andrews, B. J. (2009). Reporter-based synthetic
genetic array analysis: a functional genomics approach for investigating the cell cycle
in Saccharomyces cerevisiae. Methods Mol Biol 548, 55-73.
Schena, M., Shalon, D., Davis, R. W., and Brown, P. O. (1995). Quantitative monitoring of
gene expression patterns with a complementary DNA microarray. Science 270, 467-
470.
Schlabach, M. R., Luo, J., Solimini, N. L., Hu, G., Xu, Q., Li, M. Z., Zhao, Z.,
Smogorzewska, A., Sowa, M. E., Ang, X. L., et al. (2008). Cancer proliferation gene
discovery through functional genomics. Science 319, 620-624.
Schuldiner, M., Collins, S. R., Thompson, N. J., Denic, V., Bhamidipati, A., Punna, T.,
Ihmels, J., Andrews, B., Boone, C., Greenblatt, J. F., et al. (2005). Exploration of the
function and organization of the yeast early secretory pathway through an epistatic
miniarray profile. Cell 123, 507-519.
Schwabish, M. A., and Struhl, K. (2006). Asf1 mediates histone eviction and deposition
during elongation by RNA polymerase II. Mol Cell 22, 415-422.
Schwabish, M. A., and Struhl, K. (2007). The Swi/Snf complex is important for histone
eviction during transcriptional activation and RNA polymerase II elongation in vivo.
Mol Cell Biol 27, 6987-6995.
Shalon, D., Smith, S. J., and Brown, P. O. (1996). A DNA microarray system for analyzing
complex DNA samples using two-color fluorescent probe hybridization. Genome Res
6, 639-645.
148
Shaner, N. C., Campbell, R. E., Steinbach, P. A., Giepmans, B. N., Palmer, A. E., and Tsien,
R. Y. (2004). Improved monomeric red, orange and yellow fluorescent proteins
derived from Discosoma sp. red fluorescent protein. Nat Biotechnol 22, 1567-1572.
Shaner, N. C., Patterson, G. H., and Davidson, M. W. (2007). Advances in fluorescent
protein technology. J Cell Sci 120, 4247-4260.
Shaner, N. C., Steinbach, P. A., and Tsien, R. Y. (2005). A guide to choosing
fluorescent proteins. Nat Methods 2, 905-909.
Sikorski, R. S., and Hieter, P. (1989). A system of shuttle vectors and yeast host strains
designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics
122, 19-27.
Silva, J. M., Marran, K., Parker, J. S., Silva, J., Golding, M., Schlabach, M. R., Elledge, S. J.,
Hannon, G. J., and Chang, K. (2008). Profiling essential genes in human mammary
cells by multiplex RNAi screening. Science 319, 617-620.
Simon, I., Barnett, J., Hannett, N., Harbison, C. T., Rinaldi, N. J., Volkert, T. L., Wyrick, J.
J., Zeitlinger, J., Gifford, D. K., Jaakkola, T. S., and Young, R. A. (2001). Serial
regulation of transcriptional regulators in the yeast cell cycle. Cell 106, 697-708.
Solomon, M. J., Larsen, P. L., and Varshavsky, A. (1988). Mapping protein-DNA
interactions in vivo with formaldehyde: evidence that histone H4 is retained on a
highly transcribed gene. Cell 53, 937-947.
Solomon, M. J., and Varshavsky, A. (1985). Formaldehyde-mediated DNA-protein
crosslinking: a probe for in vivo chromatin structures. Proc Natl Acad Sci U S A 82,
6470-6474.
149
Sopko, R., Huang, D., Preston, N., Chua, G., Papp, B., Kafadar, K., Snyder, M., Oliver, S.
G., Cyert, M., Hughes, T. R., et al. (2006). Mapping pathways and phenotypes by
systematic gene overexpression. Mol Cell 21, 319-330.
Spellman, P. T., Sherlock, G., Zhang, M. Q., Iyer, V. R., Anders, K., Eisen, M. B., Brown, P.
O., Botstein, D., and Futcher, B. (1998). Comprehensive identification of cell cycle-
regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization.
Mol Biol Cell 9, 3273-3297.
Stern, M., Jensen, R., and Herskowitz, I. (1984). Five SWI genes are required for expression
of the HO gene in yeast. J Mol Biol 178, 853-868.
Sternberg, P. W., Stern, M. J., Clark, I., and Herskowitz, I. (1987). Activation of the yeast
HO gene by release from multiple negative controls. Cell 48, 567-577.
Sutton, A., Bucaria, J., Osley, M. A., and Sternglanz, R. (2001). Yeast ASF1 protein is
required for cell cycle regulation of histone gene transcription. Genetics 158, 587-
596.
Tackett, A. J., Dilworth, D. J., Davey, M. J., O'Donnell, M., Aitchison, J. D., Rout, M. P.,
and Chait, B. T. (2005). Proteomic and genomic characterization of chromatin
complexes at a boundary. J Cell Biol 169, 35-47.
Tait-Kamradt, A. G., Turner, K. J., Kramer, R. A., Elliott, Q. D., Bostian, S. J., Thill, G. P.,
Rogers, D. T., and Bostian, K. A. (1986). Reciprocal regulation of the tandemly
duplicated PHO5/PHO3 gene cluster within the acid phosphatase multigene family of
Saccharomyces cerevisiae. Mol Cell Biol 6, 1855-1865.
Timmons, L., and Fire, A. (1998). Specific interference by ingested dsRNA. Nature 395,
854.
150
Tong, A. H., Boone, C. (2005). Synthetic genetic array (SGA) analysis in Saccharomyces
cerevisiae. In Yeast Protocols (Totowa NJ, The Humana Press Inc.), pp. 171-192.
Tong, A. H., Boone, C. (2007). High-throughput strain construction and systematic synthetic
lethal screening in Saccharomyces cerevisiae. In Yeast gene analysis - methods in
microbiology (Elsevier Ltd.), pp. 369-386, 706-707.
Tong, A. H., Evangelista, M., Parsons, A. B., Xu, H., Bader, G. D., Page, N., Robinson, M.,
Raghibizadeh, S., Hogue, C. W., Bussey, H., et al. (2001). Systematic genetic
analysis with ordered arrays of yeast deletion mutants. Science 294, 2364-2368.
Tong, A. H., Lesage, G., Bader, G. D., Ding, H., Xu, H., Xin, X., Young, J., Berriz, G. F.,
Brost, R. L., Chang, M., et al. (2004). Global mapping of the yeast genetic interaction
network. Science 303, 808-813.
Typas, A., Nichols, R. J., Siegele, D. A., Shales, M., Collins, S. R., Lim, B., Braberg, H.,
Yamamoto, N., Takeuchi, R., Wanner, B. L., et al. (2008). High-throughput,
quantitative analyses of genetic interactions in E. coli. Nat Methods 5, 781-787.
Vilela, C., Ramirez, C. V., Linz, B., Rodrigues-Pousada, C., and McCarthy, J. E. (1999).
Post-termination ribosome interactions with the 5'UTR modulate yeast mRNA
stability. Embo J 18, 3139-3152.
Wach, A., Brachat, A., Alberti-Segui, C., Rebischung, C., and Philippsen, P. (1997).
Heterologous HIS3 marker and GFP reporter modules for PCR-targeting in
Saccharomyces cerevisiae. Yeast 13, 1065-1075.
Weeraratna, A. T. (2005). Discovering causes and cures for cancer from gene expression
analysis. Ageing Res Rev 4, 548-563.
151
Whitfield, M. L., Sherlock, G., Saldanha, A. J., Murray, J. I., Ball, C. A., Alexander, K. E.,
Matese, J. C., Perou, C. M., Hurt, M. M., Brown, P. O., and Botstein, D. (2002).
Identification of genes periodically expressed in the human cell cycle and their
expression in tumors. Mol Biol Cell 13, 1977-2000.
Williams, S. K., Truong, D., and Tyler, J. K. (2008). Acetylation in the globular core of
histone H3 on lysine-56 promotes chromatin disassembly during transcriptional
activation. Proc Natl Acad Sci U S A 105, 9000-9005.
Winzeler, E. A., Shoemaker, D. D., Astromoff, A., Liang, H., Anderson, K., Andre, B.,
Bangham, R., Benito, R., Boeke, J. D., Bussey, H., et al. (1999). Functional
characterization of the S. cerevisiae genome by gene deletion and parallel analysis.
Science 285, 901-906.
Wittenberg, C., and Reed, S. I. (2005). Cell cycle-dependent transcription in yeast:
promoters, transcription factors, and transcriptomes. Oncogene 24, 2746-2755.
Woychik, N. A., and Hampsey, M. (2002). The RNA polymerase II machinery: structure
illuminates function. Cell 108, 453-463.
Xu, F., Zhang, K., and Grunstein, M. (2005). Acetylation in histone H3 globular domain
regulates gene expression in yeast. Cell 121, 375-385.
Xu, H., Kim, U. J., Schuster, T., and Grunstein, M. (1992). Identification of a new set of cell
cycle-regulatory genes that regulate S-phase transcription of histone genes in
Saccharomyces cerevisiae. Mol Cell Biol 12, 5249-5259.
Yuan, G. C., Liu, Y. J., Dion, M. F., Slack, M. D., Wu, L. F., Altschuler, S. J., and Rando, O.
J. (2005). Genome-scale identification of nucleosome positions in S. cerevisiae.
Science 309, 626-630.
152
Zhang, Y., Moqtaderi, Z., Rattner, B. P., Euskirchen, G., Snyder, M., Kadonaga, J. T., Liu,
X. S., and Struhl, K. (2009). Intrinsic histone-DNA interactions are not the major
determinant of nucleosome positions in vivo. Nat Struct Mol Biol 16, 847-852.
Zhu, C., Byers, K. J., McCord, R. P., Shi, Z., Berger, M. F., Newburger, D. E., Saulrieta, K.,
Smith, Z., Shah, M. V., Radhakrishnan, M., et al. (2009). High-resolution DNA-
binding specificity analysis of yeast transcription factors. Genome Res 19, 556-566.