Transcript INO1

H2A.Z-Mediated Localization
of Genes at the Nuclear
Periphery Confers Epigenetic
Memory of Previous
Transcriptional State
Donna Garvey Brickner, Ivelisse Cajigas,
Yvonne Fondufe-Mittendorf, Sara Ahmed,
Pei-Chih Lee, Jonathan Widom, Jason H.
Brickner (PI)
Northwestern University
Background – localization of
chromatin at the nuclear periphery
•In metazoan cells, heterochromatin and developmentally repressed
genes localize at the periphery
•In yeast, peripheral localization of telomeres and the mating type loci
promotes silencing
•In yeast and mice, some genes are recruited to the periphery when
transcriptionally activated
•It is thought that peripheral localization may be a mechanism of
efficiently coupling transcription with export
•Periphery tethering points can serve as hard boundaries to limit the
spreading of heterochromatin formation
Background – INO1 and GAL1
•GAL1 - Galactokinase, phosphorylates alpha-D-galactose to
alpha-D-galactose-1-phosphate in the first step of galactose
catabolism
•activated by growth in galactose rather than glucose
•INO1 - encodes inositol 1-phosphate synthase, the enzyme
that catalyzes the conversion of glucose 6-phosphate to
inositol 1-phosphate
•activated by inositol starvation
•Because of their easy transcriptional activation or repression
through choice of growth medium, they are used here as
convenient reporter genes
Technique 1 – quantitative assay for
subnuclear localization of genes –
Lac tagging
•A series of lac repressor-binding sites (lac operators) are introduced
adjacent to a gene of interest by homologous recombination…
Gene ORF
L-O
L-O
L-O
L-O
•…In yeast strains expressing a GFP-tagged lac repressor,
introduced on a plasmid
LacI– LacI– LacI– LacI–
GFP GFP GFP GFP
Gene ORF
L-O
L-O
L-O
L-O
Technique 1 – quantitative assay for
subnuclear localization of genes –
Lac tagging
•For visual reference, the nuclear envelope protein Sec63 is tagged
with myc by similar molecular genetic methods
•Cells are fixed, fluorescently labelled myc antibodies are introduced,
and both these antibodies and GFP are visualized by fluorescence
microscopy
Nuclear envelope
Sec63
Sec63
myc
LacI– LacI– LacI– LacI–
GFP GFP GFP GFP
myc
Gene ORF
Sec63
Sec63
myc
myc
L-O
L-O
L-O
Sec63
L-O
myc
Question 1 – does transcriptional
activation induce periphery recruitment?
• Lac tagging method used to count the proportion of cells in which
GAL1 or INO1 localize to the periphery before and after
transcriptional activation is induced by glucose or inositol
starvation
• Dynamic range of the experiment: URA3, which is randomly
distributed in the nucleus, co-localizes with Sec63-myc in 25% of
cells – this is the minimum/baseline. INO1, artificially tethered to
the nuclear envelope, co-localizes in 80% of cells – this is the
maximum/ceiling
Cell Population
Percent of cells co-localizing with Sec63-myc
INO1 – uninduced - - - 31 +/- 1 %
INO1 – induced - - - - - 60 +/- 5 %
GAL1 – uninduced - - - 35 +/- 1 %
GAL1 – induced - - - - - 70 +/- 2.5 %
From: H2A.Z-Mediated Localization of Genes at the Nuclear Periphery Confers Epigenetic Memory of Previous Transcriptional State
Brickner DG, Cajigas I, Fondufe-Mittendorf Y, Ahmed S, Lee PC, et al. PLoS Biology
Vol. 5, No. 4, e81 doi:10.1371/journal.pbio.0050081
The proportion of
cells
demonstrating
INO1 periphery
localization
before and after
induction by
inositol
starvation
Sec63-myc, which shows the nuclear
envelope and ER, is in red. LacI-GFP, here
associating with Lac operator repeats
joined to the INO1 gene, is in green. This
demonstrates a cell that would not be
counted as demonstrating INO1 periphery
localization, and a cell that would.
Technique 2 – Real-Time PCR
A cell strain of interest is grown in the appropriate conditions,
then the cells are harvested and lysed to expose their mRNA
The mRNA is converted to cDNA by incubation with reverse
transcriptase
A set of primers appropriate to the gene of interest is
designed, and the PCR reaction is carried out in a special
machine in the presence of a fluorescent probe
standard PCR
schematic
http://www.ncbi.nlm.nih.gov/books/bv.fcgi?highlight=PCR&rid=mboc4.figgrp.1590
Technique 2 – RealTime PCR
The probe contains a
fluorescent dye along with a
quencher
The advancing polymerase,
encountering the probe,
cleaves it, separating the dye
from the quencher and
increasing the fluorescent
signal
The increase in fluorescent
signal intensity is therefore
proportional to the number of
probes bound, i.e. the number
of copies present
http://www.servicexs.com/plaatjes/TaqMan_RT-PCR_assay.jpg
If more cDNA copies are
initially present, the signal
increases at a faster rate, and
initial mRNA concentration is
thus indirectly measured
Question 2 – How fast does periphery
localization occur relative to
transcriptional activation – could one
process be driving the other?
• INO1 and GAL1 mRNA expression, following induced transcriptional
activation, was quantified at various time points using real time PCR
• ACT1, a constitutively expressed gene with fairly constant cellular
mRNA levels, was used as a control – absolute mRNA levels were
not quantified, but rather relative levels as compared with ACT1
• Peripheral localization (measured as a percentage of cells) was
assessed at various time points by the lac tagging method
• These two parameters were compared linearly, and with the mRNA
expression levels plotted on a log scale
Question 2 – How fast does periphery localization
occur relative to transcriptional activation
B: INO1 C: GAL1
Left side: Both parameters
plotted linearly
Right side: peripheral
localization plotted linearly,
mRNA expression plotted on
log scale
• A rapid increase in INO1 and
GAL1 mRNA expression was
observed following induction
by nutrient starvation
• This increase correlated with a
rapid recruitment of both gene
loci to the nuclear periphery
• The maximum rate of mRNA
accumulation for both INO1
and GAL1 occurred after
nuclear periphery localization
was largely complete
Question 3 – Do the genes remain localized to the
periphery after transcription becomes repressed?
• Repressed transcription by adding inositol to INO1-expressing cells
and glucose to GAL1-expressing cells
• Monitored mRNA expression and peripheral localization by the same
procedures as above
• mRNA expression for both strains returned to the repressed level
within 30 minutes
• Peripheral localization, however, persisted for more than two hours in
both cases
Question 3 – Do the genes remain localized to the
periphery after transcription becomes repressed?
The INO1 and GAL1
gene loci remain
localized at the nuclear
periphery more than
two hours following
transcriptional
repression, despite a
rapid return to
repressed-level mRNA
levels
• Cells are capable of rapidly changing the distribution of these loci, as
evidenced by their rapid peripheral localization following gene
activation, as well as the high diffusion constant for the repressed
GAL1 locus (reported by Cabal et. al. 2006)
• Therefore, an active mechanism must be employed to retain these
genes at the periphery
Question 4 – Is continued INO1 periphery localization
merely due to low but sufficient levels of transcription?
• INO1 may remain at the periphery because adding inositol to the
cells is not sufficient to achieve a full repression effect, and INO1 is
still being transcribed
• To test this, mutant strain rbp1-1 was used. These cells lose RNA
polymerase II function above the restrictive temperature.
• Cells starved with inositol to activate INO1 transcription and periphery
localization  cells raised to the restrictive temperature, inactivating
RNA polymerase II and completely ending INO1 transcription 
INO1 localization was monitored
Question 4 – Is continued INO1 periphery localization
merely due to low but sufficient levels of transcription?
Left panel: INO1 peripheral
localization at restrictive temperature
Black dots: wild type strain
Black squares: temperature sensitive
RNA polymerase mutant
White circles: mutant grown in
repressed conditions
Right panel: INO1 mRNA levels
predictably fall after raising the cells
to the restrictive temperature and
inactivating RNA polymerase II
Conclusion: Once activated,
continued transcription is not required
to maintain INO1 at the nuclear
periphery
Question 5 – Is actual transcription required to establish
INO1 periphery recruitment?
• Next, it was tested whether active transcription by RNA polymerase
II, in addition to inositol starvation, is required for INO1 to attain
periphery localization
• To test this, the rbp1-1 mutant was
– Raised to the restrictive temperature, inactivating RNA polymerase II
– Then starved of inositol
Despite being raised to the restrictive
temperature prior to induction, the rpb1-1
mutant exhibited INO1 gene localization
at the nuclear periphery to the same
extent as in the wild type strain.
•These results indicate that the
mechanism responsible for sensing
inositol starvation and recruiting INO1 to
the nuclear periphery is independent of
the activity of RNA polymerase II and the
transcriptional machinery
• INO1 and GAL1, two commonly used reporter genes, are recruited to
the nuclear periphery following the introduction of the cells to
activating conditions
• Periphery localization and increase in mRNA expression are
concurrent, with the highest level of increase in mRNA expression
occurring after periphery localization is complete
• These genes remain at the periphery long after the cells are returned
to repressing conditions and mRNA expression returns to repressed
levels
• Transcription by RNA polymerase II is not required either to recruit
INO1 to the nuclear periphery or to retain it there
Question 6 – Is INO1 and GAL1 periphery localization
inherited?
• To determine whether gene localization at the periphery is a heritable (and
thus, likely, epigenetic based) state, cells were
– Activated
– Repressed
– Maintained in logarithmic growth by dilution with growth medium
• After transcription is repressed,
INO1 and GAL1 are maintained
at the nuclear periphery, and this
state is inherited for several
generations
– 3-4 generations for INO1
– 65 generations for GAL1 (!)
• This suggests that localization is
maintained through epigenetic
mechanisms
Question 7 – Does the peripheral localization of shortterm repressed INO1 and GAL1 facilitate more rapid
reactivation?
Short-term repressed cell
populations experience much
more rapid reactivation of
GAL1 mRNA expression than
long-term repressed
populations
Short term repressed:
Galactose  glucose  galactose
Long term repressed:
Glucose  galactose
Growths carried out for several cell
generations in each medium
Question 7 – Does the peripheral localization of shortterm repressed INO1 and GAL1 facilitate more rapid
reactivation?
Short-term repressed cells experience slower
reactivation of INO1 mRNA expression
following the shift to activating conditions, due
to the specifics of activation and the
molecules involved …
…However, mutants lacking a nuclear
pore complex component necessary for
peripheral tethering experience even
slower reactivation, suggesting that
peripheral localization again has a
positive effect on reactivation rapidity
Question 8 – Can nuclear periphery localization alone
promote rapid transcriptional activation regardless of
transcriptional history?
To answer this question, a cell line was created in which the
INO1-lac operator array was supplemented with a nuclear
envelope-targetting motif called FFAT
•Tethering was found to have no
effect on steady-state levels of
INO1 mRNA expression in any
growth medium conditions
•However, artificially tethered
INO1 experienced a more rapid
increase in mRNA expression
upon inositol starvation than
untethered (wild-type) INO1
• INO1 and GAL1, two commonly used reporter genes, are recruited to
the nuclear periphery following the introduction of the cells to
activating conditions
• Periphery localization and increase in mRNA expression are
concurrent, with the highest level of increase in mRNA expression
occurring after periphery localization is complete
• These genes remain at the periphery long after the cells are returned
to repressing conditions and mRNA expression returns to repressed
levels, and this localization is inherited for several generations
• Transcription by RNA polymerase II is not required either to recruit
INO1 to the nuclear periphery or to retain it there
• The peripheral localization of short-term repressed genes facilitates
more rapid transcriptional activation
• This localization alone is sufficient to promote rapid activation,
regardless of transcriptional history
Technique 3 – Nuclease
digestion of chromatin to
reveal nucleosomal
positioning
• Cells are permeablized and treated with
nuclease to digest unprotected DNA
• RT Q-PCR and many primer sets are
used to determine relative
concentrations of overlapping DNA
sequences
• High concentrations indicate extensive
nucleosomal protection
• Well-known nucleosomal and nonnucleosomal sites are used as controls
Question 9 – Is there a difference between short-term
and long-term repressed INO1 genes in terms of
nucleosomal positioning or nucleosomal protection?
•Two stably positioned
nucleosomes identified in INO1 –
one at the promoter, and one
between the promoter and the
ORF
•There was no change in the
positioning of these nucleosomes
between short-term and long-term
repressed INO1
•There was, however, a decrease in nucleosomal protection in
short-term repressed INO1
•Decreased protection can indicate either an increased in the
proportion of cells lacking nucleosomes, or a decrease in
nucleosomal stability…and because of something about the
positioning of the two nuclesomes…they must contain…
Jason Brickner, as portrayed
by Edward Norton
H2A.Z
• A nucleosome containing a variant histone component encoded by
the htz1 gene
• Has been shown to form boundaries preventing the spread of
heterochromatin
• Inclusion at gene promoters negatively correlates with transcriptional
frequency
• Htz1 is deposited at promoters during repression, but is not required
to establish the repressed state
• It is thought that Htz1 inclusion facilitates rapid activation because
Htz1 containing nucleosomes are less stable and more easily ejected
from promoter regions than their H2A-containing counterparts
• In an in vitro chromatin system, far lower ionic strength was required
to separate Htz1 from DNA than to separate H2A or H3
Question 10 - is INO1 or GAL1 transcriptional activation
altered in htzΔ1 mutants?
Loss of H2A.Z
led to a
significant
delay in shortterm repressed
Ino1 and Gal1
reactivation
Loss of H2A.Z
had no effect
on long-term
repressed Ino1
and Gal1
reactivation
Question 11 – Are there
more H2A.Z nucleosomes
associated with short-term
repressed INO1
promoters than with longterm repressed
promoters?
Chromatin immunoprecipitation,
using an antibody that recognizes
htz1, was used to precipitate H2A.Z
and analyze the DNA sequences to
which it is bound
In long-term repressed or
transcriptionally active cells, the
bound DNA was not enriched for
the INO1 promoter
In short-term repressed cells, there
was a strong enrichment for the
INO1 promoter
This suggests that H2A.Z is
incorporated into the INO1
promoter region following
repression
Question 12 – Are htz1Δ cells deficient in peripheral
localization of the Ino1 gene?
htz1Δ cells are not deficient
in localizing Ino1 to the
periphery when actively
transcribing…
…but they are deficient in
maintaining Ino1 at the periphery
following repression
• H2A.Z nucleosomes are incorporated into the INO1 promoter
following repression
• htz1Δ mutant cells exhibit slower reactivation of short-term repressed
Ino1 and Gal1 but are unaffected as far as reactivation of these
genes after long-term repression
• htz1Δ mutant cells are able to recruit Ino1 to the nuclear periphery
following activation, but are unable to retain the gene at the periphery
following repression
Gene repression
H2A.Z incorporation
Peripheral localization
Rapid reactivation following short-term repression
The correlations are striking, but causal relationships are unclear
• Short-term and long-term repressed INO1 and GAL1 are
distinguished by their localization, transcriptional history, and kinetics
of activation
• Peripheral localization of INO1 is necessary and sufficient to promote
more rapid activation, regardless of transcriptional history
• H2A.Z incorporation is necessary to promote more rapid activation
• Unanswered question:
– Is the easy dissociation of H2A.Z nucleosomes the mechanism, or is
H2A.Z important only in promoting peripheral tethering, which itself
provides the optimal environment for reactivation?
– I would have liked to see an experiment in which genes in htz1Δ mutants
were artificially tethered to the periphery
Recruitment to the periphery
promotes H2A.Z incorporation,
which allows for rapid reactivation
due to the ability of H2A.Z to easily
dissociate from DNA
Incorporation of H2A.Z
nucleosomes promotes peripheral
localization, and the periphery
provides an optimal environment
for rapid gene activation
Significance of this work
•Demonstrates a case in which cells inherit not a transcriptional
state, the usual epigenetic paradigm, but rather transcriptional
information
•Demonstrates a novel role of H2A.Z, that of ‘marking’ recently
repressed genes and promoting their peripheral retention and
rapid reactivation
Hypothesis on why short-term
repressed INO1 takes longer to
reactivate, in case anyone asks
•Phosphatidic acid is combined with inositol to produce
phosphatidylinositol, and is the activating molecule for INO1
•Low inositol concentration leads to a buildup of phosphatidic acid, which
is responsible for the activation of INO1
•When cells are starved of inositol to cause expression, Ino1p is produced
which in turn produces inositol
•When cells are then grown in inositol to initiate repression, a significant
amount of Ino1p remains in the cell, which continues to produce inositol
and perpetuates the activity of the pathway that depletes phosphatidic acid
•This produces a lag time before enough phosphatidic acid can
accumulate to activate Ino1 expression again