Aberrant Water Structure Dynamics in B16 Melanoma-Bearing Mice by Time Domain Refractometry Analysis

Round 1

Reviewer 1 Report

The work described here is important and valuable, but unfortunately the presentation of results leaves a lot to be desired and requires further work. It is hard to get a clear understanding of whether or not tumour tissue can be distinguished from normal tissue from dispersion curves, which is the main aim of the analyses. Some statements in the discussion seem not to be accurate.

Apparently, the main comparison seems to be the pooled data from either 3 organs plus 2 regions of skin or just the 3 organs (Figure 5), and it is not made clear why these two forms of pooling are needed. It would be expected that significant differences would be found between tumour values and other organs, but the greater interest is whether a difference can be detected from measurements of organs or skin alone. Here the results are a bit inconsistent, with significant differences in beta values on day 11 but on day 14 only if the skin data is included. Although this is mentioned in lines 260 – 263, the statement ‘explicitly higher’ for TO compared to HO in line 273 is not the case for delta epsilon, only for beta. It is really important to be clear on this, because as mentioned above, HO/TO differences from other organs are intriguing. The issue of multiple comparisons (line 186) is very important and further details should be provided in the discussion section (e.g. normally the effects of multiple comparison adjustment, such as Bonferroni, is to change the critical P from 0.05 to 0.05/n, where n is the number of comparisons) . Here, the critical P seems to be unadjusted from 0.05.

1.       The headings of the results section present conclusions rather than descriptions. Conclusions are out of place in a Results section. It might be more helpful to use questions as sections headings e.g. How does tumour volume and mass vary during observation period?

2.       Actually, heading to 3.5 (line 243) is not true, because for beta there was no significant difference between TO and tumour and between either HO and TO and tumour if skins are omitted. Please amend. As mentioned, the more interesting differences are between HO and TO. Maybe be better to have questions rather than possibly misleading conclusions as sections headings (see above).

3.       Figure 3 presents ‘representative’ profiles, which implies that in order to estimate tau, delta epsilon and beta some averaging took place. The caption to figure 5 seems to imply that for each tissue 5 determinations were make (5 tissues x 5 determinations = 25). Please clarify and discuss the % error in the individual determinations. The derivation of spectra from TDR is described and referenced, but not the method of curve-fitting to estimate the parameters mentioned. For example, the usual method for determining beta is from a Cole-Cole impedance locus plot, determining the depression of the circle arc below the real axis. Was this the method followed here (this method readily gives a % error).

4.       There are several on-line resources for values of tau and beta for various tissues (for example, https://itis.swiss/virtual-population/tissue-properties/database/tissue-frequency-chart/ or http://niremf.ifac.cnr.it/docs/DIELECTRIC/AppendixC.html#Top_of_Page ). How well do your values agree with these literature values (note these have 4 Debye terms rather than the 2 used here)? Note, in particular the sigma(DC) values in the table.

5.       The caption to Figure 1 seems to need attention: (c) is the protocol and (a) is the ?scheme. Also a bit confusing to have A – E in (a) and A, 2 Bs, C & D in (b) which are undefined. Please resolve.

6.       Figure 4 presents tau versus beta data, but it is unclear how Water Structure Dynamics are revealed by behind doing these plots and whether anything useful comes of doing it. Water is an outlier as far as beta is concerned, so not clear what the justification is for the heading (line 223). It might be helpful to briefly summarise the content of ref 10 in relation to water structure, particularly in relation to transitions between bound and free water structures. The question of why the presence of tumours somewhere in the body should affect (specifically) beta values in skin and other organs generally, needs further discussion.

7.       Further to this, the title has ‘Aberrant water structure’ as a main theme, but there is no clear definition of what is ‘normal’ and what is ‘aberrant’. Please add this to the introduction, with clear definitions.

8.       The pdf supplied was hard to read because comments still remained in the RH margin. In resubmission, please ensure that these comments are removed.

There are quite a few typos (for example line 105 ‘open-ended’; contacted with what?. Line 129 ‘on ice’ – assume sample did not in fact freeze, which would affect dielectric properties markedly. Figure S3 DC in fact represents the factor sigma/(omega x epsilon0) not the DC electrical conductivity. Please check carefully through M/S.

Author Response

Dear Reviewer1

Thank you for your crucial comments.

Please see the attachment pdf files with the response to your comments and the revised manuscript.

(The revised parts have been indicated with red letters.)

Please check them.

Again, thank you for sharing your precious time.

Sincerely,

Haruchika Masuda

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript under consideration presents a dielectric spectroscopy study of tumor tissues. The authors apply time domain reflectometry to examine the differences in water structure dynamics between ex-vivo samples taken from healthy (control) mice and B16 melanoma bearing mice.

It was found that tumor samples exhibited aberrant response in comparison to organs and tissues of control animals. According to the results, the approach used can be a promising tool for diagnosing various pathophysiological conditions in vivo, including cancer.

I have some minor comments:

1) According to the Supplementary data (Fig.S3) the low-frequency component of Cole-Cole equation seems to be negligible with respect to high-frequency component. Further analysis of relaxation times, strengths, and time dispersions seems to be carried out for the latter component. I suggest to make this more clear in the text.

2) In addition to existing limitations and future experiments in Sec.4.5, I would suggest to discuss an opportunity of examining samples at extended timescale of tumor growth (e.g. 7-20 days after inoculation).

Missing comparison with histological analysis is a limitation of the current experiment. This may give more clear insights on the mechanisms of abnormal WSD in tumors in future experiments.

It would be also interesting to examine whether TDR method is able to detect the pathophysiological status of tumor after treatment with different agents (e.g. cytostatic drugs, radiation, etc).

3) Please, correct a typo in line 176 (MHz is missing).



The manuscript can be accepted for publication after minor revision.

Author Response

Dear Reviewer2

Thank you for your crucial comments.

Please see the attachment pdf files with the response to your comments and the revised manuscript.

(The revised parts have been indicated with red letters.)

Please check them.

Again, thank you for sharing your precious time.

Sincerely,

Haruchika Masuda

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have responded adequately to my concerns. The only remaining comment I have is in relation to the bottom table on page 4 of the response letter. The values for conductivity around 1 GHz seem mainly to have been copied from the permittivity table and make no sense. The left hand boxplot on p5 seem to be the experimental values, compared against the literature values. There appears to be a difference of around 3-fold between them, which may be due to species differences, but I am not convinced. Since some of the literature values are from the group represented by reference [15], it might be wise to make some comment on the differences in the actual paper.

The method of curve fitting to the Cole-Cole model is still not well explained, but I am not insisting on further changes.

Nothing further to add.

Author Response

Dear the Reviewer1,

Thank you for your crucial comments.

I have tried to response to them one by one.

Could you please see and check the uploaded pdf file as the response letter?

Sincerely, 

Haruchika Masuda

Author Response File: Author Response.pdf