2015/05/08 (Thu) MS submitted to Research Topic “Oral Oncology”
2015/05/30 (Sat) MS returned, interactive reviews activated:
Editor: Rui Amaral Mendes Catholic University of Portugal; Editor of J Carcinogenesis and Mutagenesis; Adhuant Prof at CWRU
2015/08/28 rejected… by Specialty Chief Editor JeanPascal Machiels
Reviewer#1:
加上morphology圖於Fig 1B
band quantitation
spheroid invasion
Reviewer#2:
移入IHC圖於Fig 5E
補上shRNA sequences and offtarget exp result 於Fig S1B
2015/07/13
Language:
Fine, but there are still numerous errors also in the amended sections.
E.g. Line 356: Interestingly
Line 379: only increment of DDR1 ???
Line 470: cancer
There are more! I just don’t have the time to write them down all!
Use of a spell checker and proof-reading by a native English speaker is strongly recommended.
Objective errors:
Response to #1: The standardization of the cell culture medium could confound the experiment. The change of the cell culture medium to DMEM 10% FBS could alter transcriptional programs in cell lines that are normally grown in a different medium. The resulting PTK profiles are therefore flawed.
Response to #2: I still think that the choice of reference cDNA confounds the study. To identify a role in cancer biology, the expression profiles of normal tissue must be compared with that of cancerous tissue. That means a cDNA from normal keratinocytes should have been used as a reference. Comparing a mixed population of expression profiles from unrelated cancer cell lines with the profiles of OSCC cancer cell lines is like comparing apples with pears. The compared cell lines are of different tissue origin and may therefore express DDR1 at completely different levels. These comparisons do therefore not indicate a biological significance of the identified PTKs in oral cancer.
The authors comment below that EGFR expression is even higher in normal human keratinocytes and suggest that this may be caused by continuous EGF supplementation in the medium.
These statements clearly demonstrate that DDR1 could actually be strongly expressed in normal keratinocytes and is therefore not overexpressed in OSCC lines. Furthermore, this study demonstrates the limited applicability of cell line models, especially if grown under different conditions.
Response to #4: The HPV immortalised keratinocyte cell line shows indeed reduced DDR1 levels. However, a normal keratinocyte cell line such as HaCat or HOK is still not shown.
Response to #5: The explanation that DDR1 is the only overexpressed PTK in the analysed OSCC line is based on the assumption that the profiling results are correct. As I pointed out above (Response to #2), these data are flawed.
Response to #6: ANOVA analysis was performed but I’m surprised to see statiscial significance when the means are so close and the error bars overlap.
Introduction:
Lines 110-124: Good attempt but too long. No need to summarise all results.
M&M:
Response to #1: Satisfactory. As mentioned above, a table listing the characteristics of the used cell lines would have been clearer and more appropriate. (Lines 128-142: Content satisfactory. For clarity, a table would have been more suitable.
Response to #2: The technical issues have been sufficiently addressed. However, as outlined above, the profiling data are flawed due to use of an inappropriate control.
Response to 1: The standardization of the cell culture medium could confound the experiment. The change of the cell culture medium to DMEM 10% FBS could alter transcriptional programs in cell lines that are normally grown in a different medium. The resulting PTK profiles are therefore flawed.
I still think that the choice of reference cDNA confounds the study. To identify a role in cancer biology, the expression profiles of normal tissue must be compared with that of cancerous tissue. That means a cDNA from normal keratinocytes should have been used as a reference. Comparing a mixed population of expression profiles from unrelated cancer cell lines with the profiles of OSCC cancer cell lines is like comparing apples with pears. The compared cell lines are of different tissue origin and may therefore express DDR1 at completely different levels. These comparisons do therefore not indicate a biological significance of the identified PTKs in oral cancer.
The new evidence provided by the authors demonstrate that DDR1 could actually be strongly expressed in normal keratinocytes and is therefore not overexpressed in OSCC lines. I can also see no difference between normal controls and OSCC tissue (Fig. 5E). In summary, this study demonstrates the limited applicability of cell line models, especially if grown under different conditions.
Response to 2: The DDR1 levels are lower in the HPV-immortalised cell line but not in DOK. A different control keratinocyte cell line such as OKF6 or HOK should have been included.
Response to 3: has not been addressed by the authors
Response to 10 and 11: The authors do not describe the results of all tested cell lines (only for TW2.6). If there was no effect on cell migration in these cell lines , it should be explained!
Response to 15: Not adressed by the authors.
Response to 17: Unfortunately, the images are worse than the previous version. In the previous version, membranous staining for DDR1 was evident at least in the OSCC sample. What is presented here looks mostly cytoplasmic and there is a very strong background staining. I am concerned about immunostaining of this quality could be scored accurately. The staining pattern has also not been described in the results section.
I still think that the Discussion section lacks substance.