EMBO Scientific Publishing

Thomas Lemberger

Chief Editor, Molecular Systems Biology

Deputy Head of Publications, EMBO

1. Editorial Process

2. Preparing a manuscript

3. Scientific integrity

Scientific

publishing

“The publication of scientific

information is intended to move

science forward. More

specifically, the act of publishing

is a quid pro quo in which

authors receive credit and

acknowledgment in exchange

for disclosure of their scientific

findings.”

disclose findings

credit

move science forward

critical evaluation

critical evaluation

editorial process

editorial process

@EMBO

EMBO Scientific Publications

• All areas of molecular &

cell biology

• First journal launched by

EMBO (1982)

•All areas of molecular &

cell biology

•Short-format papers

•Science & Society

section

•At the interface between

basic and clinical life

sciences

•Open Access

•Systems biology,

synthetic biology,

systems medicine

•Open Access

quality

quality

community

R Aebersold

GM Church

L Hood

E Liu

P Bork

Julie Ahringer

Charles Auffray

Ewan Birney

Tom Blundell

Thomas S. Deisboeck

Jan Ellenberg

Michael Elowitz

Alan Fersht

Stan Fields

Mark Gerstein

Frank Holstege

Sung Hou Kim

Hiroaki Kitano

Doron Lancet

Andrew J. Link

Stephen Oliver

Jeremy Nicholson

Bernhard Palsson

Rama Ranganathan

Uwe Sauer

Luis Serrano

Lucy Shapiro

Pamela Silver

Michael Snyder

Janet Thornton

Masaru Tomita

Marc Vidal

Hans V. Westerhoff

Lothar Willmitzer

John Yates

Senior Editors Advisory Board

EMBO Editors:

Thomas Lemberger

Andrew Hufton

OPEN

ACCESS

EMBO Editors:

Thomas Lemberger

Maria Polychronidou

R Aebersold

GM Church

L Hood

E Liu

P Bork

Julie Ahringer

Charles Auffray

Ewan Birney

Tom Blundell

Thomas S. Deisboeck

Jan Ellenberg

Michael Elowitz

Alan Fersht

Stan Fields

Mark Gerstein

Frank Holstege

Sung Hou Kim

Hiroaki Kitano

Doron Lancet

Andrew J. Link

Stephen Oliver

Jeremy Nicholson

Bernhard Palsson

Rama Ranganathan

Uwe Sauer

Luis Serrano

Lucy Shapiro

Pamela Silver

Michael Snyder

Janet Thornton

Masaru Tomita

Marc Vidal

Hans V. Westerhoff

Lothar Willmitzer

John Yates

Senior Editors Advisory Board

Scope & general

policies

OPEN

ACCESS

The editorial process

editorialrejection

review

reject revise

reject accept

time

First editorial decision

editorialrejection

review

reject revise

reject accept

time

First editorial decision

editorialrejection

review

reject revise

reject accept

time

First editorial decision

EMBO editors read the entire manuscript (yes!)

Decision on a balance of multiple factors:

• Scope

• Novelty & conceptual advance

• Mechanistic, functional, biological insights

• Utility of methods, dataset, resource

• Completeness and conclusiveness of the analysis

In case of doubt...R Aebersold

GM Church

L Hood

E Liu

P Bork

Julie Ahringer

Charles Auffray

Ewan Birney

Tom Blundell

Thomas S. Deisboeck

Jan Ellenberg

Michael Elowitz

Alan Fersht

Stan Fields

Mark Gerstein

Frank Holstege

Sung Hou Kim

Hiroaki Kitano

Doron Lancet

Andrew J. Link

Stephen Oliver

Jeremy Nicholson

Bernhard Palsson

Rama Ranganathan

Uwe Sauer

Luis Serrano

Lucy Shapiro

Pamela Silver

Michael Snyder

Janet Thornton

Masaru Tomita

Marc Vidal

Hans V. Westerhoff

Lothar Willmitzer

John Yates

Senior Editors Advisory Board

EMBO Editors:

Thomas Lemberger

Maria Polychronidou

Initial editorial decision

To review or not to review...

editorialrejection

review

reject revise

reject accept

time

Peer-review

Referees are invited based on:

• Balance of expertise

• Reputation as researcher

• Reputation as reviewer

• No conflict of interest (positive or negative)

3 (4) reviewers / manuscript

1. Summary

• Describe your understanding of the story

• What are the key conclusions: findings and concepts

• What are the methodology and model system

2. General remarks

• Are you convinced of the key conclusions?

• Place the work in its context.

• What is the nature of the advance (conceptual, technical, clinical)?

• How important is the advance as compared to previous knowledge?

• What audience will be interested in this?

3. Major points

• Specific criticisms related to key conclusions

• Specify experiments or analyses required to demonstrate the conclusions

• Motivate your critique with relevant citations and argumentation

4. Minor points

• Easily addressable points

• Presentation and style

• Trivial mistakes

Referee report

Transparent

peer review

http://msb.embopress.org/content/9/1/704.r

eviewer-comments.pdf

Background of this study:

Extracellular inputs are often encoded into dynamics of transcriptional factors, which are subsequently decoded into target

expression. Hao and O'Shea (2012) showed that the yeast general stress regulator Msn2 exhibits distinct shuttling dynamics

between cytoplasm and nucleus in response to different stresses. A technique based on analog-sensitive kinase (also used in current

manuscript) was developed to control the shuttling of Msn2. This technique allowed them to reveal the dependence of gene

expression on promoter kinetics under oscillatory Msn2 at the population level. In their later study, Hao and O'Shea (2013) continued

to show that the dynamics of Msn2 are largely dependent on, and can be modulated by, the phosphorylation states of the PKA target

residues on the NLS/NES of Msn2.

Short summary of this manuscript:

In this manuscript, Hansen and O'Shea took a step further and used the aforementioned technique to study the dependence of gene

expression on the dynamics of artificially generated Msn2 localization bursts in single cells. Using time-lapse microscopy and

mathematical modeling, they identified two types of promoters: high amplitude threshold slow activation (HS) and low amplitude

threshold fast activation (LF). A model with three promoter states was used to fit the experimental data and was able to capture the

experimental responses. Furthermore, they showed that the noise level in promoter expression depends on the timescale of

promoter activation as well as the transcription factor dynamics (i.e., repeated pulses vs. single pulse). Notably, slow promoter

activation is largely due to slow nucleosome remodeling. These results paint a picture of how dynamic Msn2 inputs are decoded at

the promoter level and how promoter characteristics influence the fidelity of signal decoding.

Overall comment:

These studies were carefully designed and nicely carried out. The data are solid and the analysis is comprehensive. The results

address several issues in the field of dynamic signal processing in cells and bring up interesting proposals regarding the role of

promoter kinetics and cellular noise in the decoding of dynamic signals. I enjoy reading this manuscript. I would recommend the

acceptance of this manuscript if the authors can address the following concerns.

Major comments:

1. The potential problems of the technique using PKA analog sensitive mutant....

2. The Msn2-mediated indirect regulation of target promoters....

3. The three-state promoter description...

4. The role of transcript half-life in the decoding....

5. The physiological relevance of synthetic Msn2 dynamics....

6. The chromatin remodeling timescale....

Good

Reviewer #2:

Note that I have provided a Word file that includes some figures I have generated to better show the problem with this manuscript.

The manuscript by Maria Lluch-Senar and colleagues is a bold effort to determine what are the essential genetic elements in a near

minimal bacterium. Over the last 6 years or so, this team has published a dazzling series of papers characterizing the atypical bacterium,

Mycoplasma pneumoniae, from a systems biology perspective. Because this organism, at only 860 kb and 800 genes, is already close to

being a minimal bacterial cell, determination of what genes and regions of genes are essential and what is non-essential is of great

interest. Since the days of the Max Delbruck's phage school in the 1930's it has been the dream of biologists to use a reductionist

approach to understanding how life works by identifying and determining how each essential element in a living cell functions. The aim of

this project is to identify, and to some extent characterize all the essential genetic material in a cell by marrying proteomics, and

transcriptomics data with the knowledge of what genetic material can be disrupted using transposon bombardment. The strategy

employed by Lluch-Senar and colleagues is much like the approach taken by the team of Stanford microbiologist Lucy Shapiro to identify

the essential genes in Caulobacter crescentus in 2011 that was published in Molecular Systems Biology. The Stanford team combined

hyper-saturated transposon mutagenesis coupled with high-throughput sequencing. The protein coding genes, non-coding RNAs and

other un-translated genetic elements such as origins of replication and transcriptional control regions that were not hit by transposons

were defined as essential.

In an absolute sense, transposon bombardment only defines the non-essential regions of the genome. Even then, it is possible if the

experiment is not set up properly to incorrectly label an essential gene as non-essential. For instance, a gene encoding a protein for

which there are many copies in a cell, but for which only one copy is essential could be disrupted by a transposon and still result in cells

that might survive 10 or more cell divisions. Thus, after the transposon reaction it is necessary to culture cells long enough that the entire

pool of proteins initially present in transposon disrupted cell is exhausted. Similarly, as reported in both the Stanford Caulobacter paper

and the Barcelona Mycoplasma manuscript being reviewed, disruption of some genes results in cells that have impaired growth

phenotypes, called Fitness genes. To determine the essential regions of a genome one must make several assumptions about the

transposon bombardment process:

• Transposon insertions are completely random

• The number of independent transposon mutants being analyzed is great enough that there is a high probability that all non-essential

targets are hit.

The Stanford Caulobacter project used a pool of ~800 thousand viable Tn5 transposon insertion mutants. Tn5 transposition is almost

random as was demonstrated by Fred Blattner and colleagues (Systematic Mutagenesis of the Escherichia col Genome. 2004. Journal of

Bacteriology 186(15):4921.

In the EMBL/CRG Mycoplasma pneumoniae paper under review the story is quite different. First, Tn4001 insertions are not random, as

was shown in other studies using Tn4001 in Mycoplasma genitalium and Mycoplasma pneumoniae (Glass et al. 2006, and Hutchison et

al 1999). Some genes were hit many more times than others. In Glass et al. and also in French et al.'s study of essentiality of

Mycoplasma pulmonis, several thousand individual mutants were isolated and the locations of each transposon insertion were

determined. In many cases, for example with M. genitalium gene MG_414, several hundred of the ~4000 Tn4001 mutants analyzed had

the transposon insertion at the same base of that gene. Second, and more important, not enough mutants were analyzed to make

accurate predictions about what genes, especially small genes, are essential. The EMBL/CRG Mycoplasma pneumoniae paper stated

that they obtained a pre-existing pool of 2976 Tn4001 mutants first reported on in 2006 and later in 2007 by Halbedel and Stulke. That

said, consider the likelihood a non-essential genetic element will be disrupted.

A 2007 paper by Halbedel and Stulke, which was offered by the EMBL/CRG Mycoplasma pneumoniae paper authors cited to explain the

probability of identifying specific fractions of the total set of non-essential genes and other genetic elements offers this formula: The

probability that a mutation in a gene of interest comprised of g base pairs will be found in a given collection of transposon mutants. Here n

is the number of mutants, l is the non-essential genome size, and P is the probability.

n = log (1-P) / log(1-(g/l))

It is assumed that given that all genes in M. genitalium have orthologous genes in M. pneumoniae, and that the genes present in M.

pneumoniae but absent in M. genitalium are non-essential. Thus the 236 kb of DNA in M. pneumoniae (816 kb) not in M. genitalium (580

kb) are non-essential. Additionally, Glass et al. disrupted another 100 M. genitalium genes (comprising 112 kb) and others have found

another 8 kb of non-essential M. genitalium genes. So the non-essential M. pneumoniae genome is at least 356 kb. Assuming the

minimum small RNA coding gene one is searching for is 100 bp, then one has only a 57% chance of disrupting that gene. If the minimum

target is a 100 amino acid protein, then there is a 92% chance it will be disrupted. If the minimum target were a 100 base pair gene or

noncoding RNA then the probability is on a bit over 50% it would be disrupted and categorized as not essential.

About that same point, the authors state: "With this number of mutants, the probability of finding a desired mutant in the library is

99.999%." This is almost the same sentence in the 2006 paper by Halbedel et al. where they calculated the probability of finding a mutant

among their pool of 2976 mutants that had a disrupted hprK gene. The Halbedel sentence was: "With this number of mutants, a hprK

mutant is included in the library with a probability of 99.999%." The hprK ORF is 936 nucleotides in length. By my calculation, using

current reports of the non-essential genome in M. genitalium which indicates the non-essential genome is 356 Kb, the probability today

Reviewer #2:

The manuscript by Maria Lluch-Senar and colleagues is a bold

effort to determine what are the essential genetic elements in a

near minimal bacterium. [...] The protein coding genes, non-

coding RNAs and other un-translated genetic elements such as

origins of replication and transcriptional control regions that were

not hit by transposons were defined as essential.

[…]

The probability that a mutation in a gene of interest comprised of

g base pairs will be found in a given collection of transposon

mutants. Here n is the number of mutants, l is the non-essential

genome size, and P is the probability.

n = log (1-P) / log(1-(g/l))

[…]

Quite simply, the conclusions reached by the EMBL/CRG group

about essential small ORFs and non-coding RNAs are not

supported given they only analyzed a few less than 3000

mutants. They probably should have analyzed at least 100

thousand clones, and probably several times that amount.

Importantly, they should do mutagenesis with either Tn5 or

Tn4001, plate 100s of thousands of mutants. […]. I would think

this could all be done in a month is MiSeq instead of HiSeq

sequencing was used.

This paper is altogether dispensable. I

doubt that the paper would even qualify for

specialist journals.

The short study uses an E. coli host to screen metagenomic DNA in the

pursuit of DNA encoding functions that afford tolerance to a number of

otherwise inhibitory compounds. Such compounds are claimed to be

bottlenecks for the metabolic engineering of E. coli for biomass conversion.

Authors do find some DNA clones of environmental origin that help E. coli to

survive better in the presence of such compounds. Analysis of the

corresponding sequences reveals that they include metabolic genes, recA and

membrane proteins. Such a relatively simple exercise is presented in an

overdone synthetic biology jargon.

1. The use of E. coli as a host for capturing biological functions in

metagenomic DNA libraries has been in the literature for quite a long time.

2. The work is claimed to be oriented towards improving the performance of E.

coli for biomass conversion. But there there is not any follow up. Authors show

no connections between the environmental genes identified and any

improvement of E. coli.

3. Inhibitory compounds derived from lignocellulose are likely to appear

simultaneously in a given process. There is no evidence shown that the

combination of the tolerance functions identified in the metagenome results in

a better endurance when facing a mixture of inhibitors.

4. The genes found in the selection look like quite "housekeeping". This

suggests that tolerance to the toxic compounds under study results from minor

changes in existing functions rather than appearance of dedicated resistance

mechanisms.

This is a fascinating paper in which the XX lab presents a novel platform for

discovering effective enzymes for biomass conversion. This work is based on

extracting metagenomic DNA from arbitrary environmental sources. This work

is novel, interesting and timely, given growing interest in synthetic biology and

bioenergy. I have only a couple suggestions for enhancing the presentation of

the paper and its potential impact:

1. The text is quite brief and terse. I recommend that the authors expand the

text somewhat, particularly explaining how the basis of the platform and how it

can be implemented.

2. The authors should discuss other applications for the platform. Could it be

used for example to discover more effective antimicrobial peptides? How

about uncovering enzymes for other applications such as the biosynthesis of

polymers and other materials?

This is an exciting paper which presents a

novel, significant platform. I recommend it

for publication

Bad

Cross-commenting

reviewersreports

authors

“…[it] settles a controversy in the field

which has been going on for more than ten

years. In summary, this is a landmark paper.

I cannot support publication in your journal

strongly enough!”

“As written, the paper is focused on the

methods, which is fine given that's where it

makes its most substantial contribution. But

the writing is quite technical and could

benefit from more explanation of the high-

level logic of their approach.”

“After reading through this nicely-executed

technical work, one is left with an

impression that after all we really have not

gained much new mechanistic insights.

…in addition to the XXX data that should be

generated under their current framework…”

…Each reviewer has numerous suggestions

about how to do this. It will likely be

impossible to incorporate them all while

retaining a coherent narrative. […]

Reviewer 3 also calls for an additional

experiment - including XXX stains in the

current dataset. To incorporate this into

their current analytical framework, the

authors would have to find parameters and

reagents to allow simultaneous imaging of 5

genes (not just the 4 presented

here). Moreover, they would then have to

reacquire all images using the 5-stain

protocol.

While I agree that it would be useful to

have XXX data included, I also believe that

this is beyond the scope of this paper.

Re

f #

1R

ef

#2

Re

f #

3

Ref #2 cross-comments

Cross-commenting

Revision

• Remember that the aim of peer review is to make the paper better!

• Write a detailed point-by-point response to all the referees’ issues. Make it as easy

as possible for the editor/referees to evaluate your modifications.

• Highlighting the changes in the text can be helpful.

• Attempt to address all major issues with substantial revision before re-submitting.

• In case you disagree with the suggested experiments or when in doubt, contact the

editor.

• Remember there’s a next time, and there are other journals in the same league.

Author response:

example

Post-revision review

editorialrejection

review

reject revise

reject accept

time