SCI经验之坚持到底

从06年底开始写英文文章，反反复复折腾了3个多月，完成初稿。篇幅比较长，改得自己都山穷水尽，没有耐心再看一眼了，投了个当时4.6的杂志，想试试吧，一个多月后，收到回复，果不其然，几乎被reviewer批得体无完肤。从写作基本的拼写，语法，到实验方法，统计处理，结果描述...当然，总体评价还算过的去。最主要是因为写作水平导致的表达和沟通不畅。自己根据要求，改了改后。Boss觉得还有戏，找了一个A国的师兄和他的Boss帮忙。再投，再拒，而且总体评价还不如第一次的好，有要求补实验的意思。我哪肯，本来已经不抱希望准备改投其它杂志了，还是有点不甘心，再修并据理力争，最后兵不血刃被接收了。

平时在论坛吸收不少养分，今天把两次的decision letter和我的回复(R)全部奉上，算是也回馈一点吧，希望对初投者有所帮助。letter比较长（因为我毛病多啊^o^），见笑了，大家各取所需吧。

June 11, 2007

Editorial Team:

Comments to the Author:

This is a comprehensive study on the cellular basis of tolerance induction in liver transplantation. It has to be emphasized that the concept of tolerance due to Kupffer cell activity is not new. Similar experiments have been performed in the 1980s (Caller MP et al., Transplantation 1989). The role of FasL was first introduced into this concept by the landmark study of Sun Z et al. (Liver Transplant 2003). The current paper confirms earlier reports and provides more detail on the mechanisms involved. It contains loads of data and should be seriously taken into account for publication in Liver Transplantation. However, currently the manuscript is not in an appropriate state for publication.

Reviewer: 1

Comments to the Author

This is a comprehensive study on the cellular basis of tolerance in liver transplantation. The sheer amount of data included is very impressive though not completely novel (Caller MP et al., Transplantation 1989; Sun Z et al., Liver Transplant 2003).

MINOR POINTS

- The paper should be thoroughly reviewed by a native speaker. There are a number of typos and grammatical errors sometimes leading to mistakable content.

R: I have to apologize that there are many typos and grammatical errors in my manuscript. I am not good at English writing, although it takes me a lot of time to accomplish it. The comments of the two reviewers energize me to emphasize the language learning. We have to ask a native English speaker for writing assistance.

- The survival curves nicely show differences between GdCl3 and controls. However, no information is provided regarding the causes of death in the animals. Autopsy data should be included.

R: We think the cause of death should be an acute rejection of liver graft. We do twenty cause of liver transplantation in each group in part I, twelve to do some experiments and the rest eight to observe survival time. In survival observation, rats became depressed, inactive, little eating and dark yellow urine. Autopsy found the liver became slight enlargement and mottled brown. And acute rejection evidence was shown with the light and electron microscope and liver function test in experiment rats (Fig. 2, Tab. 2).

- Results section 4.2 gives an impression on the presence of acute rejection in GdCl3-treated donor livers. What was the frequency of rejection compared to controls? Please include data on the staging of rejection. In my opinion, the data on survival and rejection are central and should be as detailed as possible.

R: In the survival observation, we found rats pretreatment with GdCl3 died almost on 2 or 3th day post transplantation and histological examination confirmed the acute rejection. We are also curious about it, so we choose the 0, 24, 48, 72 hours after transplantation as the time point to explore role of KC in this process in part II of our in vitro experiments. And the results (e.g. quantity of CTL, apoptosis of CTL, fas/fasL expression…) showed: it changed at 24h, highlighted at 48h and moderated at 72h. In part III of the further in vivo experiments, 0 and 48 hours after transplantation were chosen as the time point to verify our hypothesis.

- Legibility of some figures needs attention (e.g. Figure 4).

R: The contrast and brightness of some figures is not optimal and it has been adjusted.

- It would have been preferable to use real time PCR for quantification of mRNAs.

R: To be honest, we do not think about the adoption of real time PCR in the planning stage of the studies. Reverse transcription polymerase reaction (RT-PCR) was performed in our previous studies. We are skilled to use it and the results are stable. I agree with your suggestion and we will try it in the further studies.

- Figure 7.3 – In my impression the intensity of the bands on the gel does not correspond very well with the graph. Is this really a significant difference?

R: Yes, I am quite sure it is a significant difference. Legibility of some figures needs attention as you noted above, the contrast and brightness of this figure is not optimal and I am sorry that make you confused. I do not change the figure but only adjust its contrast and brightness with the Microsoft Office Word 2003.

- Figures 7.4, 7.5 – The graph erroneously shows a sign of significance.

R: I think the sign of “*” make the review confused, but I do emphasize the “*” represents p>0.05 which means no significance in the figure illustration.

Reviewer: 2

Comments to the Author

This work is novel and addresses a very important area in transplantation, that of mechanisms of immune tolerance. The scientific methodology is sound save for the choice of using a syngeneic animal model to study immune tolerance.

A major concern is the use of a liver transplantation model that is syngeneic (SD to SD) to study the process of immune tolerance which is important for graft acceptance in the setting of allogeneic transplantation. How can their results then be translated to and meaningful for the clinical arena of liver allograft transplantation?

R: In order to explore the role of KC in immune tolerance, we choose SD-SD liver transplant model because it has the spontaneous tolerance without any immunosuppression. We could easily and purely observe KC and its function in this process. If we use the fully allogenic Lewis to BN rat liver transplantation (a classic acute rejection model) to explore effect of KC, immunosupressants must be used to control the acute rejection. Immunosupressants have an extensive effect on immune regulation, we feel it is difficult to understand whether the blockage of KC (e.g. GdCl3) or the immunosupressants play a major role in immune tolerance.

Other concerns:

1)  There are numerous grammatical errors throughout the manuscript hence, I recommend that, in the future, the authors have a thorough edit of the manuscript by a native English speaker.

R: I have to apologize that there are numerous grammatical errors throughout the manuscript. I am not good at English writing, although it takes me a lot of time to accomplish it. This comment energizes me to emphasize the language learning. We have invited a native English speaker for writing assistance.

2)  In the various areas in the text and in the figure legends there are omissions of time points that are studied. For example, 4.4 page 15, and in the figure legend for 4-1 pg 36, the 0h time point is left out despite the data for that time point being sited.

R: Omission the data of 0h time point is convenient to compare the two groups in my opinion. As you mentioned, I revise the description and display the full data.

3)  The manuscript is too long and has too many tables and figures.

R: I agree with you. You know we have done a plenty of work in this study, in order to make everyone clearly and fully knows what we have done and understand the results better, I use many tables and figures to make the results concrete. Many tables can be left out and some figures (e.g. figure 2) could be replaced by word description for publication.

修改后：

October 2, 2007

Editorial Team:

Comments to the Author:

A few issues still need to be addressed and clarified: 1) The language needs a little more work 2) Controls on the protein gels (like Actin, etc) should be done; 3) The point of less FASL RNA/KC with same TNFa RNA/KC should at least be addressed: are the authors claiming that the KC themselves produce less FASL after GdCl3 treatment, OR are the authors claiming that there are fewer KC (and thus less FASL) because of GcCl3 treatment, or both?

Reviewer’s Comments to Author:

Reviewer:

Much experimental work has been done in this study, for which I congratulate the authors. Significant problems exist, however, that prevent the robust support for the above hypothesis. Namely:

1. There are problems with using GdCl3 for KC depletion: though this method has been used for many years, it is unclear if GdCl3 partially eliminates KCs, decreases the phagocytic activity of KCs, or impairs other functions of KCs. Relevant to this study: the authors seem to be confused as to what GdCl3 does to KCs. Does it decrease the number (as shown in Fig 2), decrease the phagocytic activity (as stated but not shown in Fig 2), or decrease/change the function of KC (as implied by less FasL mRNA production per KC?).

R: It is generally accepted that GdCl3 decreases/depletes KC and inhibits their function, especially their phagocytosis and superoxide production. (Hardonk, J Leukocyte Biol, 1992; Kino***a, Shock 2005. GdCl3 specifically impairs KCs phagocytosis, alloantigen processing and presentation abilities, without conspicuous effects on hepatocytes, endotheliocytes, circulating monocytes and other macrophages (Neyrinck AM, J Hepatology 2002; Ayala A, J Trauma 1997) and it has been conventionally applied in blocking KC functions in ischemia-reperfusion injury and liver transplantation experiments (Jahnke C, Transplant Proc. 2006; Giakoustidis D, Surg Today 2006; Jung JY, Arch Pharm Res 2005; Vajdova K, Liver transplantation 2004; et al). Similar to above, our results manifest that GdCl3 not only decrease KC number but also decrease KC functions, such phagocytic activity and FasL expression.

2. Why would there be less FasL mRNA made in KC, but the same amount of TNF-α (Fig 6-2 and Fig 9-3)? This implies that GdCl3 is having some specific effect on KC function (not KC depletion).

R: We have emphasized that there are some controversial issues about the effects of GdCl3 on KC in the discussion. It has also been suggested that two morphological subpopulations of KC exist, although there seems has little effective method to distinguish them. One type of KC dominantly produces superoxide, and another type of KC produces proinflammatory cytokines in endotoxemia. These two types of KC may be different in the sensitivity to GdCl3. GdCl3 stimulates KC to increase TNF-α production and then induces liver injury and mouse mortality after LPS challenge despite the fact that KC greatly decreased in number (Kino***a, Shock 2005). Our results were partly similar to it. A further study is needed to elucidate whether GdCl3 directly augments the function of TNF-α producing KC or indirectly augments their function by the depletion of superoxide-producing KC, which may normally suppress the function of TNF-α producing KC.

3. Similarly, if GdCl3 is acting by depleting KC (or impairing phagocytosis), why would GdCl3-treated KC make less FasL protein than non-treated KC? (The treated and non-treated samples should, after all, be normalized for total protein). This data (Fig 6-5) implies that it is not the NUMBER of KC, but the function of the GdCl3-treated KC that accounts for the difference in FasL production.

R: As mentioned above, GdCl3 could not only decreased KC numbers, but also decreased KC functions. It is a duplicate effect. Therefore, GdCl3-treated KC make less FasL protein than non-treated KC.

4. Densitometry of RT-PCR products on agarose gels is unacceptable for quantitating RNA today, and should be done with more robust methods, like real-time PCR.

R: Reverse transcription polymerase reaction (RT-PCR) could be semi-quantitation and conventionally applied to RNA analysis in our previous studies. We are skilled to perform it and the results are stable. I agree with your suggestion and we will try it in the further studies.

5. Flow cytometry is unreliable for quantitating total cell numbers from solid tissues (due to variable loss during the isolation process). This method should not be done to determine numbers of TCs in liver. Instead, immunohistochemistry should be done for accurate quantitation. Flow cytometry in this context is useful for determining RELATIVE fractions of lymphocytes, but not TOTAL numbers.

R: Thank you for your suggestions; we will change the presentation in the result and discussion part.

Suggestions for providing robust support for the novel and important hypothesis put forward by the authors (to wit, that KC induce TC apoptosis though FasL, thus decreasing rejection in a liver transplantation model):

1. Deplete KC with liposomal clodronate; this method works much better, with sole function of KC depletion

R: I really appreciate that you give me such a great suggestion. To be honest, we do not get this idea that depletion KC with liposomal clodronate in the planning stage of this study. GdCl3 specifically impairs KCs phagocytosis, alloantigen processing and presentation abilities, without conspicuous effects on hepatocytes, endotheliocytes, circulating monocytes and other macrophages (Neyrinck AM, J Hepatology 2002; Ayala A, J Trauma 1997) and it has been conventionally applied in blocking KC functions in ischemia-reperfusion injury and liver transplantation experiments (Jahnke C, Transplant Proc. 2006; Giakoustidis D, Surg Today 2006; Jung JY, Arch Pharm Res 2005; Vajdova K, Liver transplantation 2004; et al). We also notice that clodronate is a nontoxic bisphosphonate already in use for the treatment of osteoporosis, osteolytic bone metastases, and bone resorption including Paget disease and intravenous injection of liposome-encapsulated clodronate ensures that clodronate is efficiently trapped by splenic and hepatic macrophages, killing the cells without affecting other organs (Van Rooijen N, J Immunol Methods 1994; J Leuk Biol 1997; Sturm E, J Hepatol 2005) when we review the literatures. And the depletion of splenic and hepatic murine macrophages by liposome-encapsulated clodronate (lip-clod) was studied as a new strategy for immune thrombocytopenic purpura treatment (Fernanda, Blood 2000). But if we replace GdCl3 with clodronate, it will take 1 or 2 years to wholly re-perform this study. It is a novel method for us to block KC function and we are seriously interested in Liposome mediated depletion of macrophages in our further investigation.

2. Provide normalization controls for all protein quantitation studies, use real-time semi-quant PCR for mRNA quantitation.

R: The normalization control gels were added, thank you for your suggestion. Reverse transcription polymerase reaction (RT-PCR) could be semi-quantitation and conventionally applied to RNA analysis in our previous studies. We are skilled to perform it and the results are stable. I admit that it is a shortage when we plan this study, and we will try it in the further investigations.

3. To fully show cause and effect: TCs must also be depleted as a control, to ensure that the decreased survival is indeed due to TC-mediated rejection (and not a different effect of the GdCl3 treatment)

R: It is well known that TCs play a crucial role in immune modulation, contributing to graft injure and rejection. There seems little effective methods to TC depletion although many investigations focused on it. We think it is really difficult for us to make TC depleted.

终篇：

December 10, 2007

We are pleased to inform you that your revised manuscript has been accepted for publication. Your manuscript will be sent to the publisher, John Wiley & Sons. You should receive the galley proofs directly from the publisher...

写文章，其实就四个字：“坚持到底”。

与大家共勉。