Download Gene promoters dictate histone occupancy within genes

The EMBO Journal Peer Review Process File - EMBO-2013-84880
Manuscript EMBO-2013-84880
Gene promoters dictate histone occupancy within genes
Roberto Perales, Benjamin Erickson, Lian Zhang, Hyunmin Kim, Elan Valiquett and David L.
Bentley
Corresponding author: David L. Bentley, University of Colorado School of Medicine
Review timeline:
Submission date:
Editorial Decision:
Revision received:
Editorial Decision:
Accepted:
24 February 2013
02 April 2013
12 July 2013
29 July 2013
30 July 2013
Transaction Report:
(Note: With the exception of the correction of typographical or spelling errors that could be a source of ambiguity,
letters and reports are not edited. The original formatting of letters and referee reports may not be reflected in this
compilation.)
Editor: Anke Sparmann
1st Editorial Decision
02 April 2013
Thank you for submitting your research manuscript entitled "Spt6 depletion reveals that histone
occupancy within genes is promoter-dependent" (EMBOJ-2012-84280) to our editorial office. It has
now been seen by three referees and their comments are provided below.
All reviewers appreciate your study and are in general supportive of publication in The EMBO
Journal. Nevertheless, referee #1 and #3 do raise a number of concerns that should be taken into
consideration, especially the request for a clarification of the normalization procedure (Ref #1 Point
1) and the testing of additional promoters of each type in the promoter swapping experiment, ideally
with different levels of transcription (Ref #1 Point 2 & Ref #3 Point 3).
Given the comments provided, I would like to invite you to submit a suitably revised manuscript to
The EMBO Journal that attends to the raised concerns in full. I should add that it is our policy to
allow only a single major round of revision and that it is therefore important to address all criticism
at this stage. Please do not hesitate to contact me should any particular argument require further
clarification.
When preparing your letter of response to the referees' comments, please bear in mind that this will
form part of the Review Process File, and will therefore be available online to the community. For
more details on our Transparent Editorial Process, please visit our website:
http://www.nature.com/emboj/about/process.html
Thank you for the opportunity to consider your work for publication. I look forward to your
revision.
© European Molecular Biology Organization
1
The EMBO Journal Peer Review Process File - EMBO-2013-84880
-----------------------------------------------REFEREE COMMENTS
Referee #1
Perales et al. in this manuscript investigate the role of the nucleosome assembly factor Spt6 in
controlling nucleosome occupancy, particularly in transcribed regions. Previous work had shown
that Spt6 contributed to nucleosome re-assembly over coding regions of highly transcribed genes,
and that effects on nucleosome occupancy were not closely correlated with effects on transcription,
by looking at a much more limited fraction of the yeast genome. The present work confirms these
findings and goes considerably beyond them by examining H3 (and by inference nucleosome) and
pol II occupancy genome-wide in wild type and spt6 yeast, using a degron mutant, and employing
both ChIP-chip and ChIP-seq. New findings are made on the relative effects on 5' and 3' regions of
transcribed genes, on the effects on genes having different promoter types, and on an interesting
effect of promoter type on the effect of Spt6 over the adjacent coding region. Overall these findings
are of considerable interest, and I would rate the work between Medium and High in novelty and
general interest.
First, it is not clear from the somewhat abbreviated description provided in the Methods whether the
increased nucleosome occupancy reported at some regions could be due in part to normalization
methods. For instance, if normalization is such that the average signal is forced to be zero-so it is
equal overall between wild type and spt6 mutant yeast-then some regions will necessarily show
apparent increased histone occupancy in the latter cells to compensate for the regions that show
decreased occupancy. The authors could address this by performing qPCR verification comparing
some specific regions showing increased H3 occupancy to a region that is expected to be unaffected,
such as centromeric chromatin. At the very least they should explicitly address this possibility by a
better discussion of their normalization procedure.
Second, the effect of different promoter type on the effect of the spt6 mutant on histone occupancy
in the adjoining transcribed region is very interesting, but (I almost hate to say this) only one
example of each promoter type is used to replace the YLR454W promoter for this experiment. If
more than one example of each were shown it would strengthen the result-but I am hesitant to say
this is really needed.
Other minor points:
The authors use "chromatin repair" in a couple of places to refer to reassembly of nucleosomes (e.g.
in Abstract, "that repairs transcribed chromatin"). This should be avoided, as it leads the reader (at
least this one) to think of transcription-coupled repair of DNA damage, quite a different
phenomenon.
Supplemental Figure 2e needs a label (rRNA) for the graph.
What are the extra dashed curves in Fig. 1b?
Kim et al. (2010) referred to in the text (e.g. Fig. 1 legend) was published in 2011 according to the
bibliography.
Figure 1c and f are very close to redundant-they show the same data graphed relative to the starting
ATG and to the TSS. One could be eliminated or at least moved to supplemental data. In contrast,
there is definite value to separately showing ChIP-chip and ChIP-seq data to corroborate each other.
There is an error in the Fig. 2 legend "(groups 1-3 n=). Also, I don't find any indication for colors in
2b (they are mentioned in the text but not the figure legend), and short and long genes don't appear
to be marked in 2c, though the legend makes it sound like they should be.
The authors remark on an observed difference between pol II ChIP and NRO-seq data, particularly
for RP genes. Similar observations were made by Pelechano et al. (V. Pelechano et al. 2009 PLoS
Genetics 8: e1000614), who reported a higher fraction of apparently inactive pol II associated with
RP genes than other genes, so that they showed higher ChIP for Pol II than indicated in
corresponding GRO-seq data.
© European Molecular Biology Organization
2
The EMBO Journal Peer Review Process File - EMBO-2013-84880
What are the red bars in Figs. 2d-i?
Are DPN/OPN lines in Supp Fig 5B reversed? It looks like there is a broader open region in OPN,
while it should be opposite.
Fig. 6a-identify wt and spt6.
Fig. 6b and d-labeling (position wrt TSS?) is not indicated; is "GCD1-YLR" referring to TSS?
p. 13, "At ADH1, H3 was strongly depleted throughout the gene (Fig. 6d)" - should be 6a
Figs. 6c and e; don't mix technical replicates with biological replicates. The figure indicates n=6 but
the legend states that the number of biological replicates is at least three for each IP. Averages and
associated error bars should be reported using only biological replicate data; first average technical
replicate (PCR values) for each IP, then average and calculate errors using n= # of biological
replicates.
Referee #2
This is a high quality and very solid analysis of the role of Spt6 on transcription and nucleosome
occupancy globally in budding yeast. The unexpected findings are that Spt6, a chromatin assembly
factor promotes maintenance/formation of the nucleosome depleted regions at promoters. Also, the
drastic loss of nucleosomes downstream of the TSS in the Ty elements upon spt6 inactivation is
striking. I see no deficiencies in the paper as it stands.
Referee #3
This manuscript presents data characterising the changes in the organisation of the transcription and
chromatin apparatus following depletion of the histone chaperone Spt6. This analysis is carried out
to a high standard. The conclusions to be drawn are quite closely related to previous studies of the
effects of an Spt6 ts allele. However, the final couple of figures present data indicating that genes
with different classes of promoter architecture respond differently to depletion of Spt6. This
suggests that chromatin may be organised differently over genes that are transcribed in bursts in
comparison to genes that are transcribed more continuously. Additional support for this is gained
from a promoter swapping experiment. This latter part of the manuscript represents something that
is so far as I am aware a new concept that could be a significant general interest.
Minor points:
1) P11"On the other hand, after Spt6 inactivation, there is little difference in average pol II
occupancy
between "fast" and "slow" genes (Fig. 4e, f, green traces). In summary, these results
suggest that Spt6-mediated histone replacement is a major determinant of nucleosome
occupancy within genes with high histone exchange rates."
Another interpretation of the data would be that RNA polymerase is the major determinant of
nucleosome occupancy at fast genes.
2) YLR454W belongs to the closed class, yet H3 chip is little altered following Spt6 degron. The
change looks more similar to that occurring at the GCD1 promoter which belongs to the open class.
There should be some discussion of this.
3) Transcription when driven by GCD1 is 2-fold lower than when driven by ADH1.
The latter point is acknowledged as a limitation of this approach in the text. Really the most rigorous
way to deal with this would be to repeat the experiment using different promoters from the open and
© European Molecular Biology Organization
3
The EMBO Journal Peer Review Process File - EMBO-2013-84880
closed classes that are transcribed at different levels. This is tedious but I think necessary to
convince sceptics.
1st Revision - authors' response
12 July 2013
Perales et al Point by point response:
-----------------------------------------------Referee #1
Perales et al. in this manuscript investigate the role of the nucleosome assembly
factor Spt6 in controlling nucleosome occupancy, particularly in transcribed
regions. Previous work had shown that Spt6 contributed to nucleosome reassembly
over coding regions of highly transcribed genes, and that effects on
nucleosome occupancy were not closely correlated with effects on transcription,
by looking at a much more limited fraction of the yeast genome. The present
work confirms these findings and goes considerably beyond them by examining
H3 (and by inference nucleosome) and pol II occupancy genome-wide in wild
type and spt6 yeast, using a degron mutant, and employing both ChIP-chip and
ChIP-seq. New findings are made on the relative effects on 5' and 3' regions of
transcribed genes, on the effects on genes having different promoter types, and
on an interesting effect of promoter type on the effect of Spt6 over the adjacent
coding region. Overall these findings are of considerable
interest, and I would rate the work between Medium and High in novelty and
general interest.
First, it is not clear from the somewhat abbreviated description provided in the
Methods whether the increased nucleosome occupancy reported at some
regions could be due in part to normalization methods. For instance, if
normalization is such that the average signal is forced to be zero-so it is equal
overall between wild type and spt6 mutant yeast-then some regions will
necessarily show apparent increased histone occupancy in the latter cells to
compensate for the regions that show decreased occupancy.
Thanks for raising this point which is a very valid one as in ChIP-Chip experiments
log2 ratios of ChIP signal/input are scaled to be centered around zero and ChIP-seq
results are normalized by calculating reads per bin per million mapped reads (RPBM)
and we have noted this in the revised methods section (p. 20, 21). As a result these
experiments reveal relative changes in histone occupancy not absolute changes. We
have modified our conclusions to reflect this limitation. We note however that elevated
H3 ChIP signals in spt6 relative to WT as measured by Q-PCR at some amplicons (Fig.
6c, E, Supp. Fig. 6E) are consistent with net histone deposition at some locations when
Spt6 is inactivated.
The authors could address this by performing qPCR verification comparing some
specific regions showing increased H3 occupancy to a region that is expected to
be unaffected, such as centromeric chromatin. At the very least they should
explicitly address this possibility by a better discussion of their normalization
procedure.
Second, the effect of different promoter type on the effect of the spt6 mutant on
histone occupancy in the adjoining transcribed region is very interesting, but (I
almost hate to say this) only one example of each promoter type is used to
replace the YLR454W promoter for this experiment. If more than one example of
each were shown it would strengthen the result-but I am hesitant to say this is
really needed.
We made additional strains in isogenic WT and spt6-td backgrounds (DBY1392, 1396,
1398, 1402 Table 1) with YLR454W fusions to another OPN promoter (PDC1) and
© European Molecular Biology Organization
4
The EMBO Journal Peer Review Process File - EMBO-2013-84880
another DPN promoter (APA1). Results with these strains shown in Supp. Fig. 6 and
discussed on p. 13 are in good agreement with the GCD1 and ADH1 fusion genes in
our original submission.
Other minor points:
The authors use "chromatin repair" in a couple of places to refer to reassembly of
nucleosomes (e.g. in Abstract, "that repairs transcribed chromatin"). This should
be avoided, as it leads the reader (at least this one) to think of transcriptioncoupled
repair of DNA damage, quite a different phenomenon.
Done.
Supplemental Figure 2e needs a label (rRNA) for the graph.
Thanks, done.
What are the extra dashed curves in Fig. 1b?
The dashed curves delimit the range of values that includes the central 80% of genes.
Kim et al. (2010) referred to in the text (e.g. Fig. 1 legend) was published in 2011
according to the bibliography.
Thanks for catching this omission. The missing reference has been added.
Figure 1c and f are very close to redundant-they show the same data graphed
relative to the starting ATG and to the TSS. One could be eliminated or at least
moved to supplemental data. In contrast, there is definite value to separately
showing ChIP-chip and ChIP-seq data to corroborate each other.
We have deleted the redundant Fig. 1f and replaced it with corroborating ChIP-ChIP
data for short, medium and long genes formerly in a supplemental figure.
There is an error in the Fig. 2 legend "(groups 1-3 n=). Also, I don't find any
indication for colors in 2b (they are mentioned in the text but not the figure
legend), and short and long genes don't appear to be marked in 2c, though the
legend makes it sound like they should be.
The figure and legend have been corrected. We have also modified Fig. 2b to include
data from a new pol II ChIP-seq data set from W303 WT cells that has greater
dynamic range than the previous ChIP-ChIP data.
The authors remark on an observed difference between pol II ChIP and NRO-seq
data, particularly for RP genes. Similar observations were made by Pelechano et
al. (V. Pelechano et al. 2009 PLoS Genetics 8: e1000614), who reported a higher
fraction of apparently inactive pol II associated with RP genes than other genes,
so that they showed higher ChIP for Pol II than indicated in corresponding GROseq
data.
We have included acknowledgment of the Pelechano et al paper on p. 8. Note however
that this discrepancy between ChIP and NRO-seq is somewhat reduced when we used
ChIP-seq rather than ChIP-ChIP data to calculate pol II density in the revised Fig. 2b
probably because ChIP-seq is more accurate.
What are the red bars in Figs. 2d-i?
Figure corrected
Are DPN/OPN lines in Supp Fig 5B reversed? It looks like there is a broader
open region in OPN, while it should be opposite.
Supp. Figs 5a, b have been revised using ChIP-seq data.
Fig. 6a-identify wt and spt6.
Figure corrected.
Fig. 6b and d-labeling (position wrt TSS?) is not indicated; is "GCD1-YLR"
© European Molecular Biology Organization
5
The EMBO Journal Peer Review Process File - EMBO-2013-84880
referring to TSS?
Figure legend clarified.
p. 13, "At ADH1, H3 was strongly depleted throughout the gene (Fig. 6d)" should be 6a
Corrected.
Figs. 6c and e; don't mix technical replicates with biological replicates. The figure
indicates n=6 but the legend states that the number of biological replicates is at
least three for each IP. Averages and associated error bars should be reported
using only biological replicate data; first average technical replicate (PCR values)
for each IP, then average and calculate errors using n= # of biological replicates.
Corrected
Referee #2
This is a high quality and very solid analysis of the role of Spt6 on transcription
and nucleosome occupancy globally in budding yeast. The unexpected findings
are that Spt6, a chromatin assembly factor promotes maintenance/formation of
the nucleosome depleted regions at promoters. Also, the drastic loss of
nucleosomes downstream of the TSS in the Ty elements upon spt6 inactivation is
striking. I see no deficiencies in the paper as it stands.
Referee #3
This manuscript presents data characterising the changes in the organisation of
the transcription and chromatin apparatus following depletion of the histone
chaperone Spt6. This analysis is carried out to a high standard. The conclusions
to be drawn are quite closely related to previous studies of the effects of an Spt6
ts allele. However, the final couple of figures present data indicating that genes
with different classes of promoter architecture respond differently to depletion of
Spt6. This suggests that chromatin may be organised differently over genes that
are transcribed in bursts in comparison to genes that are transcribed more
continuously. Additional support for this is gained from a promoter swapping
experiment. This latter part of the manuscript represents something that is so far
as I am aware a new concept that could be a significant general interest.
Minor points:
1) P11"On the other hand, after Spt6 inactivation, there is little difference in
average pol II occupancy between "fast" and "slow" genes (Fig. 4e, f, green traces). In summary,
these
results suggest that Spt6-mediated histone replacement is a major determinant of
nucleosome occupancy within genes with high histone exchange rates."
Another interpretation of the data would be that RNA polymerase is the major
determinant of nucleosome occupancy at fast genes.
Our results establish the particular importance of Spt6 for nucleosome maintenance on
“fast” exchanging genes. They in no way eliminate the possibility that RNA
polymerase is a major determinant of nucleosome displacement from these genes. We
have changed the text on p. 11 to reflect this point. “In summary, these results suggest
that Spt6-mediated histone replacement is a major determinant of nucleosome
occupancy within genes with high histone exchange rates where active displacement
correlates with relatively high pol II occupancy.”
2) YLR454W belongs to the closed class, yet H3 chip is little altered following
Spt6 degron. The change looks more similar to that occurring at the GCD1
promoter which belongs to the open class. There should be some discussion of
this.
We have mentioned on p. 12 that YLR454W is an atypical member of the “closed” class
in this respect.
© European Molecular Biology Organization
6
The EMBO Journal Peer Review Process File - EMBO-2013-84880
3) Transcription when driven by GCD1 is 2-fold lower than when driven by ADH1.
The latter point is acknowledged as a limitation of this approach in the text.
Really the most rigorous way to deal with this would be to repeat the experiment
using different promoters from the open and closed classes that are transcribed
at different levels. This is tedious but I think necessary to convince sceptics.
As mentioned above in response to reviewer 1, we made additional strains in the WT
and spt6-td backgrounds (DBY1392, 1396, 1398, 1402 see Table 1) with YLR454W
fusions to another OPN promoter (PDC1) and another DPN promoter (APA1). These
new results are shown in Supp. Fig. 6 and discussed on p. 13 are in good agreement
with the GCD1 and ADH1 fusion genes in our original submission.
We have also strengthened the case for promoter specific effects on histone eviction by
recalculating eviction in subsets of OPN/DPN and open/closed genes chosen on the
basis that their average pol II densities within the gene body are approximately
matched. These results in the revised Figure 5 show that greater eviction in
OPN/closed genes can not be attributed to greater pol II occupancy.
2nd Editorial Decision
29 July 2013
Thank you for submitting your revised manuscript for our consideration. It has now been seen once
more by one of the original referees, and based on her/his comments, I am happy to inform you that
we are ready to proceed with acceptance of the paper, pending modification of a few minor points.
- Please implement the changes suggested by the referee in Figure 6.
- I would like to suggest defining "Spt phenotype" as "Suppressor of Ty (Spt) phenotype" in the
abstract.
- Please complete and sign the linked license agreements (see below).
Since you currently do not have access to your manuscript through our tracking system, please
simply send the amended figure and text via email. We will then upload the information into our
system.
If you have any questions, please do not hesitate to contact me directly.
Thank you for your contribution to The EMBO Journal!
___________________________________
REFEREE COMMENTS
Referee #1
The authors have done a good job of addressing previous criticisms, and particularly the manuscript
benefits from the inclusion of additional tests of the effect of promoter type on Spt6-mediated
histone occupancy over the ORF (Fig. 6). However, the error bars in Figure 6 are still a problem.
The legend now explains that only two biological replicates were used in Figs. 6b-e, but the error
bars are smaller than in the original version in which the figure indicated that n=6 and n=4. It
appears the authors are still using these values of n, rather than the correct n=2, and have now used
standard error of the mean rather than SD. SD should be used, with n=2; or rather than a histogram
the individual data points (for biological replicate data only, gotten from the average of the technical
replicates) could be shown.
Note the reference to Pelechano et al. on p. 8 needs to be formatted.
© European Molecular Biology Organization
7
The EMBO Journal Peer Review Process File - EMBO-2013-84880
Italicize SGA1 on p. 9.
On p. 11, "DPN/opn" should be "DPN/open".
Additional correspondence (author)
30 July 2013
Many thanks for the news.
I have attached a revised Figure 6 and revised text with the abstract modified as you suggested and
the corrections
noted by referee #1. We will send you the GEO accession number for the ChIP-seq and ChIP-Chip
data sets as soon as it is assigned.
Additional correspondence (editor)
30 July 2013
Thank you for providing the amended figure and text. I am very pleased to inform you that your
manuscript has now been officially accepted for publication in the EMBO Journal.
We do have your license agreements on file, and please just add the GEO accession numbers of the
ChIP data sets at the proof stage.
Thank you for your contribution to The EMBO Journal!
© European Molecular Biology Organization
8