Dear Dr. Davis,
Thank you very much for considering our manuscript for publication in Molecular Neurobiology. I am now submitting a revised manuscript for your consideration.
Briefly, both of the reviewers requested a better English writing and re-organization of the introduction and discussion. We have carefully rewrote the parts of introduction and discussion. Since both reviewers required the transverse sections stain, we also performed immunohistology on transverse section at 4 h time point, the results are shown as supplemental figure (figure legend is in blue). Below, you will find a point-to-point response in italics to the reviewers’ comments. I hope that the considerable amendments will make it possible for you to accept our manuscript for publication.
We hope that by responding to all reviewers’ comments and adding new experiments our study will be acceptable for publication
Sincerely yours,
YanQin Shen
> Reviewer #1: There are a number of points that are inconsistent thru the text/figures and better insight into the roles of HMGB1 may be deduced with further experimentation- particularly to examine additional time points and more strongly test the roles of HMGB1 at various time points and with more mechanistic insights.
Answer: We verified the changes in HMGB1 during the acute and chronic stages after SCI. The 4 h and 12 h time points represent the acute stage, and the 6 and 11 d time points represent the recovery and chronic stages. Furthermore, we studied 21 d, which is a more chronic stage of regeneration. Since it is very difficult to observe the HMGB1 in adult SCI zebrafish by live imaging, we examined these 5 time points to adequately cover the duration of recovery. For the function of HMGB in regeneration, we found downregulation of HMGB1 inhibits both locomotor recovery and the promotion of axonal regrowth after SCI. We also provide evidence for the ability of HMGB1 to promote angiogenesis after SCI, which is consistent with the beneficial role of HMGB1 on locomotor recovery. We agree that more mechanisms of HMGB1 during regeneration should be studied in detail in the future. However, specific mechanistic studies are beyond the scope of the paper.
> Reviewer #2: In this study, the authors characterized the expression of HMGB1 in the adult zebrafish for sham and spinal cord transected animals, and assessed the effect of HMGB1 on functional recovery. The zebrafish is an excellent model system to study spinal cord injury and one of the corresponding author's laboratory has long history of excellent work in this field. The current study is certainly of interest to the field and can provide useful insight into HMGB1's contribution to recovery from spinal cord injury. Especially intriguing is the decreased expression during recovery and increased association of neurons with blood vessels after injury. However, portions of this manuscript are poorly written and the study is lacking experimental elements to support their conclusions. As such, there are several major points that need to be addressed.
Answer: We appreciate your positive comments on our work. Stanley Li Lin revised the whole manuscript and rewrote some parts of the introduction and discussion, made the introduction more logical and the discussion more concise and less speculative.
> Major Points:
> 1. The authors state they are investigating HMGB1 based on its upregulation in the brainstem following spinal cord injury. The paper they cite, Ma et. al.,2012, (Cysteine- and glycine-rich protein 1a is involved in spinal cord regeneration in adult zebrafish) has no mention of HMGB1 anywhere in the paper. One can only assume that they are referring to the microarray analysis performed in the cited paper. If that is the case, they should discuss the magnitude of HMGB1 upregulation in the microarray analysis and the time course over which it occurs.
Answer: Yes, we referred the results of microarray in the cited article. We now introduce the magnitude of HMGB1 from the microarray results in the SCI zebrafish brainstem in the line 7 of the introduction.
> 2. The introduction section of this paper is not written very well and reads as though it's an assembly of statements that lack cohesiveness. For example, "HMGB1 has been found to be upregulated in rat spinal cord at 72 hrs after injury, and increased in rat brain from 3 days after ischemia (Kawabata et al., 2010; Kim et al., 2006). During early development in the mouse, HMGB1 is widely expressed in brain structures, including the hippocampal dentate gyrus, olfactory bulbs and cell lining of the telencephalic ventricles (Guazzi et al., 2003)." Here, the authors do not state that HMGB1 expression after injury is in direct contrast to what happens in zebrafish, and faili to explain why HMGB1 expression in the identified nuclei is of any importance or relevance to their study. The majority of the introduction is written in the same fashion. In spite of listing the identified role(s) of HMGB1, the authors do not indicate what type of protein HMGB1 is, or that it's primarily known to affect chromatin structure and transcription.
Answer: We re-organized the introduction, provide a rationale of why HMGB1 is important, and reorganize the background to suggest that HMGB1's neurogenic and inflammatory properties must be highly coordinated for neuroregeneration to occur. We also took your suggestion about the describing the role of HMGB1. We now provide more information about HMGB1 in the first paragraph of the introduction.
> 3. The use of transgenic animals in this study was nice to see and demonstrated the spatial location of blood vessels well. However, the authors did not mention if they generated these animals or if the animals were given to them by the lab that originally made them. If they made these animals, they will have to go into much greater detail of the transgenic process in the methods section. If not, they need to indicate how these animals were obtained.
Answer: We have added detailed information of the transgenic fish. The fish was kindly provided by Zilong Wen (Hong Kong University of Science and Technology), who obtained the fish from Weinstein lab whose reference was cited. Please find the information in the Materials and Methods, under the Animal section.
> 4. The qRT-PCR has no mention of a "no RT" control or "no template" control.
Answer: Thank you for pointing out the no template control. We did run a "no template" control in our experiment, which gave no Ct value. We have added this information in the Materials and Methods, Real-time quantitative RT-PCR section, and the appropriate section of the Results.
> 5. In the in-situ hybridization results, the authors write, "In the sham injury group (4 hrs after SCI), most of the HMGB1 mRNA-positive signals were from ependymal cells lining the central canal. At 4 hrs after SCI, HMGB1 mRNA-positive cells were enhanced along the central canal caudal to the lesion site." The signal in the sham panel is clearly more neuronal than ependymal, and then switches to primarily ependymal at 4 hrs. In addition, the authors should use arrows in some panels to identify the motor neurons, which are easily identified to those of us working with zebrafish experience, but not so for all readers.
Answer: You are correct about the in situ hybridization results in the sham and 4-hour groups. We have revised the description and also use arrows to mark the spinal cord motoneurons in the figures.
> 6. In figure 2C, the authors indicate, "In the sham injury group, HMGB1 was mainly localized on the membrane of ependymal cells lining the central canal, but was also detectable in cell nuclei. At 4 hrs after SCI, HMGB1 was not only detectable in nuclei, but had translocated into the cytoplasm." The 4h panel shows HMGB1 staining in cells located far away from the central canal, almost to the pial surface, but the authors fail to address this point. Sagittal sections alone are not convincing in this figure therefore transverse sections stained for HMGB1 are needed to clarify the staining pattern. Additionally, it's erroneous to write that HMGB1 can be localized to a membrane when it functions as a DNA binding protein or is secreted into the extracellular space.
Answer: According to your suggestion, we performed immunohistochemistry on the transverse sections for the 4h time points for both sham and injury groups. The staining pattern is very similar to that in longitudinal sections, where positive staining for HMGB1 was found not only around the central canal, but also in the white matter, as shown in the following supplemental figure. We also corrected the description of HMGB1 localization in the Results part in the manuscript.
Supplemental figure: HMGB1 protein expression in spinal cord at 4 h after SCI in adult zebrafish.
HMGB1 protein expression in the spinal cord 4 mm caudal to the lesion site was examined by immunofluorescence in transverse sections at 4 h after SCI. HMGB1-positive cells were mainly expressed around central canal and in white matter. * indicates the central canal. HMGB1 (red), nuclei (blue). Scale bar, 50 μm.
> 7. The statistical test in figure six has three independent groups and therefore requires a one-way ANOVA with post-hoc test, not a t-test.
Answer: Thank you for your comments. Yes, the statistical test here is one-way ANOVA with Tukey’s post-hoc test, not a t-test. We revised it.
> Minor points:
> 1. Although the number of animals used is indicated in figure legends, the methods section should also list the number of animals used for each assay.
Answer: We now list the number of animals for each assay in the methods part.
> 2. Spinal cord injury is abbreviated as SCI in the abstract but appears again in long-form in paragraph two of the introduction.
Answer: We replaced the full name of SCI with the abbreviation.
> 3. In the last sentence before the materials and methods section, there is a verb missing (i.e. demonstrating) between "lesion site, that regeneration".
Answer: We rewrote the introduction part there..
> 4. How were Western blots developed? Film? This was left out of the methods.
Answer: The Western blot assay was imaged by a Gel Image System. We have added more information in the methods.
> 5. No indication of dorsal and ventral on any histological images or in figure legends.
Answer: The spinal cord of zebrafish is very slim, making it very difficult to identify the dorsal or ventral direction after removal from the fish. Instead, we use the central canal to locate the position of our section. All of our histological sections contain the central canal to image the similar position.
> Reviewer #3: Spinal cord regeneration is highlighted for its broader social issues, and lower vertebrates with ability of spontaneous spinal cord regeneration following injury have become perfect models to investigate the relative mechanisms. In this study, the groups have found that the controversial protein HMGB1 was implicated in the zebrafish spinal cord regeneration by promoting locomotor recovery, accompanied by increased angiogenesis. The results are interesting, and will attract broader readers.
> The main problems:
> 1. In figure 1 and figure 2, the magnified transverse sections should be provided to confirm the localization of HMGB1 in cytoplasm or nucleic of ependymal cells. The current longitudinal sections are insufficient.
Answer: The PDF file may not illustrate the quality of figures well, in our original tiff files, one can clearly distinguish the location of HMGB1 in motoneurons. Ependymal cells are very small, and they are so sensitive, we are afraid that even transverse sections cannot illustrate well. Please refer to the above supplemental figure.
> 2. The author should further discuss why at 4 hr, the hmgb1 mRNA is elevated whereas the protein level is unchanged? Is there any other physiological roles?
Answer: We thank you for this suggestion. We now explain in the Discussion that: a) increased mRNA expression of HMGB1 only persists for a fairly short time, which may not produce enough protein to show the difference between sham injury group, b) SCI could activate mechanisms to degrade HMGB1 concomitant with the increase in HMGB1 mRNA.
> 3. The author should statistic the number of motorneurons after MO-treatment.
Answer: Thank you for the suggestion. Since the slides are already gone, we cannot perform statistics on the number of motoneurons without redoing the experiment, which would take about 2 months, exceeding the deadline for revision. However, we further show that after MO-treatment, the percent of axonal regeneration crossing the original lesion site is reduced, which reflects the function of HMGB1 on a histological level.
> The minor questions:
> 1. The sentences should be polished. For example: p5 by neuroinflammation, with the being pain ? alleviated after application of HMGB1 antibody (Shibasaki et al., 2010); HMGB1 is released to the extracellular space? and mediates--; p7, was extracted from the 4 mm piece (segments?); p11, A video camera recorded the trials from? above the tank; p12, in the regeneration of spinal cord injured adult zebrafish. p14, motoneurons) in uninjured zebrafish, sham-injured zebrafish and injured zebrafish at varying times after SCI. In the--; p14, HMGB1 is expressed in endothelial cells and associated with of? motoneurons after SCI;----.
Answer: A native English speaker, Stanley Li Lin reviewed carefully the whole manuscript.
> 2. The reference " Dong et al---" should give the page numbers.
Answer: We are thankful for this alert. When we edited this reference, it was only published online without the page numbers. Recently, it has the volume and page information, which we have updated our manuscript.
1