ブラジリアン柔術と共通するのですが，研究においてもオフェンス技術を向上させることは比較的簡単です。難しいのはディフェンス技術を向上させることです。相手がどんな技でこちらをパスしてくるのかは試合にならないとわからないので，ディフェンス力を上げていくためには，様々なアタックに対応できるスキルを学ぶ必要があります。
閑話休題。論文は自分たちのアピールしたいところをテンコ盛りで投稿します。いわば競合グループに対するオフェンスです。
まあしかし僕らの研究領域は狭く，投稿した論文の査読は，競合相手に必ずと言ってよいほど回されます。
かつて，「Cellular Microbiology」 に論文を投稿したとき，「この人にはレビューをまわさないで」みたいなところに競合グループのボスの名前を入れたら，「そこへ，しっかりレビューが回った」という経験があります。どうして解ったのかというと，大胆にもそのボスがレビューに名前を入れてきたからです。
よほどの大御所でない限り，提出された論文は，revise (リバイズ）の指示が出されます。また，ほとんどが追加実験を含む Major revision であることが多いです。査読者のたくさんのコメントにレスポンスすることは，とても疲れます。心が折れそうになります。。。しかしここが，研究者としてのディフェンス能力が問われるところでもあります。
まさに，論文を出してからが本当の戦いが始まるのです。晴れ晴れした気持ちでいられるのは，論文を出してから 3-4 週間ぐらいですかね。。。
【投稿した論文のゆくえ】
論文を投稿すると，同じ領域の研究者によって専門的視点に立った査読 (peer review) が行われます。査読をパスして雑誌への掲載が採択されることが受理 (accept) です。一方，内容が不十分である場合は，雑誌への掲載が拒否 (reject) されます。
前述したように，一発で論文が通ることはほとんどなく，査読者から修正・追加実験の指示がでる場合がデフォルトだと考えたほうが良いでしょう。査読者のコメントにしたがって論文の訂正 (revise) をおこない，修正論文を編集者に返送します。リバイズが完璧でないと再び査読にまわり，場合によってはリジェクトされます。
リバイズした論文につけるレターは Rebuttal letter (反論の手紙) と呼ばれますが，あまりいい表現ではないです。反論するというよりは，コメント一つ一つに誠意をもってレスポンスする必要があるからです。
僕が反論したことは，過去に1度だけです。
かつて，EMBO J. に編集者のレベルでリジェクトされたときに，編集者宛に Rebuttal letter を送って，「ちゃんと回してくれ」と要求したことがあります。運良くその論文は査読者に回され，追加実験をたくさんしてアクセプトされたという経緯があります。
すごく長々となりましたが，ここからが本題となります。
【再投稿時の Rebuttal letter について】
論文の再投稿時には研究者としてのデフェンス能力が試されるときです。査読者はこちらの弱点をこれでもか，と突いてきて，それに対する追加実験のデータを再投稿の論文に盛り込んでいくことで，論文のクオリティーが初回投稿時よりもグレードアップしていくのが査読システムの良いところです。
なので，査読者の皆さんに言いたいのは，
「重箱のスミを突くのがあなた方の仕事ではなく，確実に論文のクオリティーが上がるコメントだけに絞るべきである」
ということです。まあでも，査読の意味も解らない，張り切ったポスドクに論文がまわることもありますから，運もあるかと思います。
閑話休題。再投稿時の Rebuttal letter の内容によってはリジェクトされることもあるので，ここは慎重にいく必要があります。ここでも我々の Rebuttal letter を例にして，具体的に解説していきたいと思います。
 |
| 「ライフハックで雑用上等」から改変引用 |
まず，上記のように，原稿番号(manuscript number)を明記することを忘れないようにしましょう。最初の書き出しは，編集者と査読者に査読の労をねぎらって感謝の意を表するところからはじまります。いきなり反駁では相手もカチンときますからね。
このレターでは，査読者からきたメジャークレームについて追加実験をおこなったことを明記しました。また追加実験により，我々が発見したことを強調しておきました。
この追加実験は編集者も強く要求していて，逆にこの実験を入れればアクセプトするよ，とやんわり言ってきたので，レターの内容は簡潔です。
最後の段落で査読者のコメントにすべて応えたことを述べてから，査読者からきたコメントに一つ一つレスポンスしていきます。
PLOS ONEに出した論文では，２人の査読者がついて，約30のコメントに応えたので，A4で9枚のレターとなりました。コメントには全部レスポンスしなければなりません。
査読者がどうにもできない実験を要求してくる場合がありますが，「与えられた時間ではできなかった。将来的にクレーム内容は解決していきたい」とか，一応，誠実に対応する必要があります。
査読者からのコメントに対しては，Q & A 形式でレターに添付したほうが親切でしょう（↓の付録もご参照ください）。また，訂正箇所はページ数とセンテンスの行番号を明記します。これがいい加減だと訂正箇所を探しながらなので，査読者がイライラしてきてリジェクトしたい気持ちが湧き上がってくるので，注意しましょう。
査読者も同じ研究者です。それなりに忙しいので，相手の時間をできるだけ大切にするという配慮もアクセプトに繫がる重要なファクターだと思います。
忘れてはならないのは，どんなに間違ったコメントをしてきても査読者が支配権を握っているということです。なので，変にプライドが高いヒトは，この領域で生きていくのはツライと思います。。。無意味なコメントに応えるのは本当に馬鹿馬鹿しいですからね！
まあしかし僕は，平身低頭にコメントを書いていくタイプです。できない実験にも「時間がなくて，ごめんね」と謝るタイプです。こういう苦労があるからこそ，無事に論文がアクセプトされたときにはシャンパンをあけてラボの皆と乾杯します。IFが10を越えたときにはドンペリで乾杯します。しかし，まあそういう慶事はそんなにないのでポケットマネーでなんとかなってます。
【終わりに】
これでこのシリーズも最終回となりました。自分たちが可愛がって世の中に出した論文が，査読者にボコられてトボトボと戻ってくる様は，心が折れます。
いや，折れても良いんです。。。
ロッキー・バルボアのようにまた立ち上がり，強くなって，査読者と再戦する権利が与えられている限り，前進する価値があります。リジェクトされる場合もありますが，また気を取り直して別の雑誌に出せば良いのです。
英語が苦手な僕がここまでこれたので，このブログを読んでおられる皆さんは，きっと大丈夫です。これから色々あるかと思いますが，いつかは，自分なりの論文作成術が身につくはずです。
このブログを通して，皆さんが少しでも前進する勇気が湧いてきたら，僕も嬉しいです。頑張ってください☆
【付録 過去の Rebuttal letter 】
コメントが嫌になるほどテンコ盛りなので，どこかのフレーズは使えるでしょう（笑）
ご自由にお使いください。論文のアクセプトをお祈り致します。
＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝
日付
Dr. ◯◯
Academic Editor
PLoS ONE
U.S. Headquarters Public Library of Science
1160 Battery Street, Koshland Building East, Suite 100
San Francisco, CA 94111 United States
Re: Manuscript ID: PONE-D-12-XXXXX
Dear Dr. ◯◯,
Thank you very much for your e-mail and review of the manuscript (PONE-D-12-XXXXX) that we sent on January 25, 2012. We thank two reviewers for providing constructive comments regarding the improvement of the original manuscript. We have carried out additional experiments including the whole-cell proteome analysis to respond the reviewers’ major criticisms. Newly generated data reinforce the BspR function as a global gene regulator.
Here, we are sending a PDF file of our revised manuscript. All changes have been made in response to the reviewers’ suggestion, and itemized response to the individual reviewer’s comments are also attached.
Reviewer #1:
Major Concerns:
Q1. The use of the term global regulator has not been shown to be warranted in this context. Deletion of bspR was shown to affect a few known virulence factors and the iron response, but the actual role of BspR in regulatory pathways has not been defined. In order to claim BspR is a global gene regulator some more complete analysis should be performed, for example microarray. Additionally, statements about regulation of virulence factor genes should be cautious, as it is unknown if BspR directly affects virulence factor gene expression or interacts with known regulators of virulence factor genes. Although it was briefly discussed that BspR is a molecular switch for BvgAS, it is not clear what data support this claim. mRNA levels of bvgS or bvgA were not assessed in the bspR mutant.
A1. We appreciate helpful suggestion of Reviewer 1. As suggested, the sentence has been rephrased and the term “molecular switch” has been removed from the Abstract section (P3 line 2). As indicated, we have carried out the comprehensive analysis using whole-cell lysates from B. bronchiseptica wild type and ∆bspR strains. These data have inserted in the Results section as “BspR functions as a global regulator in B. bronchiseptica” (P15, lines 2 - 14), and new data for the whole-cell proteome analysis have inserted in Table2 and 3. In addition, raw data are also presented in the Table S1. Results of the comprehensive analysis directly show that various gene expressions are affected by the BspR mutation. For example, recombination-associated protein PdgC and transcriptional termination factor Rho are dramatically increased in the bspR mutant strain (Table 2). In contrast, biotin carboxyl carrier protein FabE and diaminobutyrate aminotransferase EctB are reduced in the lack of BspR (Table 3). These results strongly suggest that BspR functions as a global regulator in B. bronchiseptica.
Furthermore, the comprehensive analysis showed that the relative amount of BvgA in the ∆bspR was significantly higher that in B. bronchiseptica wild type (Table 2). These results suggest that BspR is involved in the regulation of the various gene expressions under control of the BvgAS system via the alteration of the BvgA concentration. These comments have been inserted in Discussion section (P19, line 12 - P20, line 16). Accordingly, we speculate that BspR negatively regulates btrS and other gene expressions via the alteration of the BvgA concentration, rather than the anti sigma factor for BtrS. We appreciate helpful comments of Reviewer 1.
Q2. The interactions with known sigma factors/anti-sigma factors should be thoroughly discussed, since this is the most likely mechanism of regulation.
A2. As indicated, the interaction with known sigma/anti-sigma factors has been described in the Discussion section (P18 lines 1 – 13).
Q3. The statistical analyses used are not described in the Materials and Methods section.
A3. As indicated, the statistical analysis has been described in the Materials and Methods section (P32 lines 6 – 9).
Q4. The standard bacterial nomenclature designating genes, in italics with a lower case first letter, should be used throughout. A deletion of the gene should be designated ?bspR, not ?BspR. BspR is the protein, and bacteriologists do not refer to the protein being deleted, only the gene. Thus there is no place for "?BspR" anywhere in this manuscript.
A4. As indicated, the term “∆BspR” is replaced by “∆bspR” throughout the manuscript.
Minor Concerns:
Q1. There is only the crudest of assays of the role of BspR (in a sentence like this either gene or protein may be used, although we understand that protein is the likely actor, and so preferred) in infection; a single survival curve at a single dose, and that apparently very close to the LD50 for the wild type strain. The best that can be concluded from these data is that the LD50s of wild type and ∆bspR strains are different, but it would be very difficult to argue these data prove the difference is even 2-fold. Based on other phenotypes (e.g. effector expression) the difference is likely to be 20-fold, 200-fold or greater. The opportunity to show this is missed here. This would greatly improve the paper. Even a brief discussion of what was done and how in vitro observations might explain the in vivo effect could improve this aspect. For example, FHA and ACT have been shown to be important for virulence of B. bronchiseptica and could potentially be contributing to the decrease in virulence of the ?bspR mutant, rather than BspR directly contributing to virulence.
A1. We sincerely accept comments made by Reviewer 1. As indicated, we have commented that attenuated virulence of the ∆bspR strain can be attributed to dysregulation of virulence factors, such as FhaB and CyaA in the Discussion section (P17 lines 6 – 18).
Q2. The relationship between the terms bspR and BB1639 should be clearly identified once, and (preferably only one of them) used consistently thereafter.
A2. As suggested, we have described BB1639 as bspR in the first paragraph of the Results section (P8 lines 5 – 6) and the term BB1639 has been removed from the manuscript thereafter.
Q3. The mutant described as a Type III secretion mutant should be referred to as ∆bscN, not as a mutant completely lacking the Type III secretion system.
A3. As indicated, we have described the type III secretion mutant as ∆bscN in the Materials and Methods section (P24 lines 11 – 12).
Q4. In the abstract it is mentioned that the three classical bordetellae species share the Type III Secretion System. While it is true that most genes are present for all three species, a functional TTSS has never been demonstrated for B. parapertussis.
A4. As indicated, the Abstract section has been rephrased properly as follows: These pathogenic species share a number of virulence genes, including the gene locus for the type III secretion system (T3SS) that delivers effector proteins (P2, lines 3 - 5).
Q5. In the introduction it is stated that B. parapertussis also causes whooping cough and is able to infect other animals. One lineage called B. parapertussis has been isolated from humans AND ONLY HUMANS. A second genetically distinct clade of B. parapertussis was isolated from sheep AND ONLY SHEEP. There is no evidence to suggest either naturally infects both sheep and humans.
A5. As indicated, the description of B. parapertussis in the Introduction section has been corrected as follows: B. parapertussis can be divided into two distinct lineages, B. parapertussisHU and B. parapertussisOV, which infect human and sheep, respectively [2]. B. pertussis and B. parapertussisHU are strictly human-adapted species and are the etiological agents of whooping cough (pertussis) (P4 lines 5 – 8).
Q6. Although Figure 1 is useful in presenting the iTRAQ scheme, it is not necessary to the paper.
A6. As indicated, the iTRAQ scheme has been removed from Figure 1.
Q7. P. 14 line 9 should be Fig 6. instead of Fig. 5.
A7. As indicated, the sentence has been corrected properly (P14 line 11). Current Fig. 5 corresponds to previous Fig. 6.
Q8. Statistics values on Fig. 5B are missing
A8. As indicated, the statics values have been added in Fig. 4B (corresponding to previous Fig. 5B).
Q9. Figure 7 - asterisks lack resolution to be clearly seen in figure.
A9. Fig. 7 has been removed from the manuscript in response to the suggestion made by Reviewer 2.
Q10. In the abstract and introduction, B. bronchiseptica has not been shown to be the actual progenitor of B. pertussis and B. parapertussis. It is often said that B. pertussis and B. parapertussis are thought to have evolved from a B. bronchiseptica -like progenitor, although this is only meaningful in terms of total gene content, since the former appear to have lost many genes still present in the latter. All currently circulating strain have exactly the same time interval since the last common ancestor connecting them all. Thus none are "more closely related" in evolutionary time to that ancestor.
A10. The abstract section has been rephrased as follows: Bordetella bronchiseptica is closely related with B. pertussis and B. parapertussis, the causative agents of whooping cough (P2, lines 2 - 3). The sentence “B. bronchiseptica or a B. bronchiseptica-like organism is thought to be an evolutionary progenitor of B. pertussis and B. parapertussis” has been removed from the Introduction section. We appreciate helpful suggestion of Reviewer 1.
Reviewer #2:
Q1. On Page 5, line 11, the authors list 5 known Bordetella type III secreted proteins, but on page 7 they list 6 secreted proteins (BscF was not listed before).
A1. As indicated, BscF was described in the Results section (P8 lines 6 – 9). We appreciate helpful comments of Reviewer 2.
Q2. Page 7, line 5: Please describe the T3SS mutant used. Is it
A2. As indicated, we have described the T3SS mutant as ∆bscN in the Materials and Methods section (P24 lines 11 – 12).
Q3. Figure 2B, please indicate that the dots represent amino acid sequence identity.
A3. As indicated, the dots represent amino acid sequence identity and we have described an explanation for the dots in the figure legend (P38 lines 7 – 8).
Q4. Pages 10-11, Figure 4A: The authors need to explicitly state in the legend that these infections were done at an MOI of 20. Previously published data (Reference 14) with B. bronchiseptica infection of L2 cells shows a ?90% cytotoxicity as early as 45 minutes post-infection, which initially seems to conflict with the data presented in Figure 4A. However, this discrepancy may be explained by the differences in MOI (100 vs. 20).
A4. As indicated, we have described an MOI in figure legend, and we also have stated in the Materials and Methods section as follows: Morphological changes in infected cells was monitored as previously described [15], except that cells were infected at an MOI of 20 (P30, lines 9 - 11).
Q5. The avirulence of the
A5. As indicated, we have validated the bacterial number in the lung. Mice were infected intranasally with B. bronchiseptica wild type or ∆bspR (5 x 106 CFU/mouse). To determine the amount of bacterial colonization in the lung at 2 h after infection (time 0), the whole lung was homogenized in 10 ml cold PBS. The resulting homogenates were serially diluted with cold PBS and plated on BG agar plates, and then colonies were counted to calculate the number of cfu per lung. The numbers of bacteria in the lung were 0.9 x 106 and 2.4 x 106 in wild type and ∆bspR infections, respectively. Although the degree of the bacterial number infected with ∆bspR strain was somewhat higher than that of infected with wild type, we have confirmed that the ∆bspR strain was delivered properly in the lung. We couldn’t get the exact colony counts at each time, since mice inoculated with wild type died of the infection.
Q6. Page 14, line 9, the SDS-PAGE figure is Figure 6, not Fig. 5.
A6. As indicated, the sentence has been corrected properly (P14 line 11). Current Fig. 5 corresponds to previous Fig. 6.
Q7. Page 15, line 10: Please describe the mutation for the EPEC T3SS-deficient strain.
A7. EPEC T3SS-deficient strain used in this study is the defective strain of escF that encodes a component of needle-structure. As suggested, the results of the TEM-1 translocation assay using this strain has been removed from the manuscript. Accordingly, description of the EPEC T3SS-deficient strain has also been deleted.
Q8. Page 16, lines 11-14: The authors cannot make a claim to the start Met solely on the migration size of the protein on a gel. The start Met needs to be determined by N-terminal sequencing.
A8. The nucleotide sequence of bspR was analyzed by MetaGeneAnnotator (http://metagene.cb.k.u-tokyo.ac.jp/) that is able to predict the ribosome-binding site of prokaryotic genes. Using this program, the nucleotide sequence of bspR was predicted to contain a open reading frame of 191 amino acids protein. The predicted amino acid sequence corresponds to the molecular mass of 20-kDa and this prediction is consistent with the results of the SDS-PAGE analysis of B. Bronchiseptica and B. Pertussis BspR proteins. These explanations have been inserted in the Discussion section (P16 lines 11 – 15). Determination of the actual start codon of BspR by the N-terminal sequencing is better than above program. However, it takes several months to determine the N-terminal sequencing of BspR.
Q9. Page 17, paragraph 1: Since regulation of type III secretion is tightly regulated by the BtrS sigma factor, it is imperative that the authors assess the role of BspR in the context of BtrS, in addition to BvgAS. The authors should look at the expression and regulatory effects of BspR in a
A9. We sincerely accept comments of Reviewer 2. As suggested, we carried out the GST-pull down assay to investigate the interaction between BtrS and BspR. However, we could not detect the interaction between BspR and BtrS sigma factor. In addition, BspR does not show any sequence similarity to previously described anti sigma factors or regulators. These explanations have been inserted into the Discussion section (P19 lines 1 – 4). Thus, we speculate that BspR is unlikely to be an anti-sigma factor for BspR. Instead, the whole-cell proteome analysis demonstrated that BspR is involved in the alteration of BvgA concentration. The relative amount of BvgA in ∆bspR was significantly higher that in B. bronchiseptica wild type (Table 2). Accordingly, BspR may functions as a molecular switch via alteration of the BvgA concentration. Indeed, previous report of BipA in B. pertussis indicates that the BvgA concentration has the critical role in the alteration of the Bvg-regulated phases. Similar results were also obtained by our proteome analysis (Table 2 and 3). The BipA expression was significantly decreased in the bspR mutation, where the abundance of BvgA was significantly increased.
In Bordetella species, the BvgAS system functions as a master regulator of the various gene expressions including btrS on the btr locus. Collectively, these results suggest that BspR is negatively regulates btrS and other gene expressions via the alteration of the BvgA concentration, rather than the anti sigma factor for BtrS. These comments have been inserted in the Discussion section (P19, line 5 – P20 line 16). We are constructing the btrS-deficient strain, but it takes several months.
We believe that additional experiments including whole-cell proteome analysis, GST-pull down assay, and complementation studies (P19, lines 8 - 11) reinforce the characterization of the BspR function. We appreciate helpful comments of Reviewer 2.
Q10. Page 18, line 13: The authors cannot claim that "BspR plays a key role in the expression of genes encoding iron-responsive virulence factors" because they have not conducted a direct, in-depth analysis of any iron-responsive genes. Their only evidence in the indirect link with type III secreted proteins, which could be a pleiotropic effect.
A10. As indicated, we have rephrased the sentence as follows: BspR is involved in the iron-responsive gene expression of virulence factors (P22 lines 4 – 5).
Q11. Page 19, lines 5-15: This entire line of reasoning is pointless and does not offer any explanation for the phenotypes seen.
A11. As indicated, we have removed following sentences from the Discussion part: Previously, we have reported that the type III effector BopN is involved in the persistent colonization of the lung by Bordetella [20]. The absence of iron is an environmental cue indicating entry into a host cell and is a signal to induce the distinct gene expression necessary to establish the virulent phase via BvgAS and BspR regulatory systems.
Q12. Page 23, line 12: Why did the authors choose to complement the
A12. Reviewer 2 correctly pointed out the important matter. In this study, we complemented the bspR mutant strain by the fhaB promoter-driven bspR expression vector. The fhaB promoter-driven vector produces certain amount of BspR that allows effective complementation of the bspR mutant. To exclude an artifactual effect by the the fhaB promoter, we also confirmed that ∆bspR was complemented by the bspR expression vector with its native promoter (Fig. S2). These explanations have been inserted in the Discussion section (P16, line 18 - P17, line 5).
Q13. The authors give no explanation for why BspR would be translocated into the host cell instead of just being secreted out, since BspR mediated repression of T3SS would be relieved either way? Where does BspR localize to within the host cell? Does it induce any phenotypes? If the authors make the claim that BspR gets translocated into host cells based on only one TEM-based assay, they need to be prepared to address these obvious questions with additional assays. Otherwise, don't include this data without proper validation.
A13. We sincerely accept comments of Reviewer 2. Although translocation of BspR could be detected on the TEM-based assay using EPEC model, we have not fully addressed function of BspR in the host cells. Therefore, we have removed the above data from the manuscript.
Q14. Figures 3C and 5B: Correct the spelling in the Y axis - "Relatuve" should be "Relative".
A14. As indicated, the spelling has been corrected (Fig. 2C and 4B).
Q15. Overall, this paper started well but then presented too many disparate phenotypes without assessing any of them thoroughly. There are too many holes in this manuscript to defend the authors' claim that BspR is a global regulator of T3SS, several Bordetella virulence factors and the iron response pathway.
A15. We sincerely accepted comments of Reviewer 2. We have removed the results of the BspR translocation into the host cells (previous Fig. 7) from the manuscript. Instead, we have focused on the regulatory role of BspR. For this reason, we have performed additional iTRAQ analysis using whole-cell lysates of wild type and ∆bspR to comprehensively characterize the BspR function on the Bordetella gene expression (Table 2). Additional data for the comprehensive analysis have inserted in the Results section (P15, lines 2 - 14) and Table 2, Table 3, and Table S1.
Q16. The discussion is weak.
A16. The discussion section has been thoroughly rephrased.
Figures and Table:
To improve the manuscript according to the reviewers’ comments, we have changed figure composition as described below:
Previous Fig. 1 and Fig. 7 have been removed.
Previous Fig. 2 has been moved to Fig. 1.
Previous Fig. 3 has been moved to Fig. 2.
Previous Fig. 4 has been moved to Fig. 3.
Previous Fig. 5 has been moved to Fig. 4.
Previous Fig. 6 has been moved to Fig. 5.
New Table has been inserted as Table 2.
Fig. S2, Table 3, and Table S1 have been inserted in the manuscript.
Previous Table 2 has been moved to Table 4
We believe that we have addressed reviewers’ comments and hope that the revised manuscript is now acceptable publication in PLoS ONE. Thank you for your generous consideration.
Sincerely yours,
自分の名前
所属
＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝＝
上記内容も「ライフハックで雑用上等」から，改変引用しております。羊土社さんで販売されているので，よろしくお願いします。
