Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study aims to compare the efficiency of plasmid DNA delivery via intradermal injection (id), sonoporation-enhanced (id/SP, without microbubbles), and electroporation-enhanced (id/EP) methods. The design encompasses a complete workflow from buffer optimization to immunogenicity validation. The core conclusion is clear: id/EP yields the best results, while id/SP without microbubbles is even less effective than id alone—a finding that holds reference value. The article is well-structured, and the data generally support the conclusions. However, limitations in the experimental design, along with insufficient depth in the interpretation and discussion of results, affect the generalizability and persuasiveness of the findings. Revisions are recommended.

1. The title is clear and accurate overall, but it is slightly lengthy and cumbersome, and needs to be optimized in terms of fluency and conciseness.
2. Different numbers of animals were used in different experimental series in the paper. To ensure the robustness of the conclusions, please explain the specific basis for determining the sample size of each group (n=3, 4, 5).
3. The "2" in "H2O" in Figure 1A should be marked as a subscript to conform to the standard expression of the chemical formula, and the same is true for the caption.
4. Why is the resolution of Figure 2 E too low?
5. It was observed that the 81 mmol/L group in Figure 1A showed 7 data points, while the other groups showed 8. Please check the data and state whether there are any missing data or plotting errors.
6. The color scheme, axis format, error bar representation, and other aspects of the charts and graphs in the full text are inconsistent in style, which affects the visual professionalism and coherence. It is recommended to unify the color scheme (such as using the color palette recommended by the journal) and standardize the format elements of all charts and graphs.
7. The authors explicitly did not use microbubbles, citing "increased cost and complexity." However, the vast majority of literature on efficient sonoporation relies on microbubbles to enhance cavitation. This makes the study an evaluation of a "non-standard" or "simplified" version of the sonoporation effect. The authors should add the key limiting condition of "no microbubbles" in the abstract and conclusion to avoid misleading readers' overall understanding of ultrasound perforation technology.
8. The control of variables is not rigorous enough: there are differences in DNA dosage (50μg, 100μg) and the number of injection sites (1, 2) in the experiment. Although the standardization method is used to try to correct it, it may still affect the direct comparability between different delivery methods, and the rationality of the variable setting is not clearly stated.

Author Response

Comments 1: The title is clear and accurate overall, but it is slightly lengthy and cumbersome, and needs to be optimized in terms of fluency and conciseness.

Response 1: The title has been modified to a shorter and simpler one: “Efficacy of plasmid DNA delivery into mice by intradermal injections alone and facilitated by sonoporation or electroporation”.

 

Comments 2: Different numbers of animals were used in different experimental series in the paper. To ensure the robustness of the conclusions, please explain the specific basis for determining the sample size of each group (n=3, 4, 5).

Response 2: We agree with the Reviewer #1 that the number of animals in the group is a crucial issue. In all series of experiments authors have chosen to minimize the number of animals to adhere to the 3R principles of animal research (replace, reduce, refine). In various experiments, we used different experimental units, namely regions of interest or sites in one animal, here called “skin patches” (one or two per mouse) or an individual animal.

For the ID, EP-1, EP-2 and SP series, using skin patches as experimental units was justified as we have studied the dynamics of reporter expression at the local site of id injection/electroporation/sonoporation. The id injection sites in the ID series did not overlap, so it was possible to use two id injection sites per mouse. This allowed for an increased sample size per group while minimizing the number of animals per group.

In the SP series, the sonoporation sites could overlap because the diameter of the sensor head was comparable to the size of the mouse. To avoid adverse effects due to overlapping sonoporation sites, only one skin site was sonicated in mice in this series. The IR series studied the antibody response to immunization, so the experimental unit within the group was the animal. Consequently, there was one injection/electroporation site per mouse.

In the EP series, we tested different electroporation regimen, so to exclude the impact of electroporation on the health of mice, we electroporation one site per mouse. The experiment EP-2 of EP series id injection and EP was performed at two sites to compare the results of ID and ID/EP deliveries monitored using different imaging devices (LumoTrace Fluo and IVIS).

The sample size for all series was calculated using the “resource equation” method (https://pubmed.ncbi.nlm.nih.gov/24250214/). This information has been added to the “Animals and Animal Experiments” section (page 5, inserted with tracking, shaded gray).

a) The ID series consisted of five groups (groups ID.1 to ID.5), with five mice each. Each mouse had two id injection sites (skin patches). The skin patch on the animal served as an experimental unit. Thus, each group contained 10 experimental units. In total, the ID series contained 50 experimental units. 

E = Total number of experimental units − Total number of groups = 50 – 5 = 45 > 20.

Here, we used five mice per group, with two experimental units per mouse. Since intradermal injections in low-ionic-strength phosphate buffers could result in significant variation, using 10 experimental units per group was justified.

b) The SP series consisted of four groups (groups SP.1 to SP.4), with three mice each. Each mouse had one id/SP site (skin patches). The skin patch on the animal served as an experimental unit. Thus, each group contained 3 experimental units. In total, the SP series contained 12 experimental units.

E = Total number of experimental units − Total number of groups = 12 – 3 = 9 < 10. 

This study was a pilot study involving three mice and was not repeated in larger groups of animals, as the efficacy of the sonoporation method was questionable. Therefore, we minimized the number of animals in the experimental groups in this series following the 3R principles (reduce, refine, replace).

c) The EP-1 series consisted of five groups (groups EP-1.1 to EP-1.5), with four mice each. Each mouse had one id/EP site (skin patches). The skin patch on the animal served as an experimental unit. Thus, each group contained 4 experimental units. In total, the ID series contained 20 experimental units.

E = Total number of experimental units − Total number of groups = 20 – 5 = 15, between 10 and 20. The number of mice in a group is acceptable.

The EP-2 series consisted of one group (groups EP-2), with three mice each. Each mouse had two id/EP sites (skin patches). The skin patch on the animal served as an experimental unit. In total, this group contained 6 experimental units.

d) The IR series consisted of five groups (groups IR.1 to IR.5). In groups IR.1 to IR.3 was five mice each, in group IR.4 – four mice, in IR.5 group – 3 mice. The experimental unit was an individual animal.  Thus, groups IR.1 to IR.3 contained 5 experimental units, group IR.4 – 4 experimental units, group IR.5 – 3 experimental units. In total, the IR series contained 22 experimental units.

E = Total number of experimental units − Total number of groups = 22 – 5 = 17, between 10 and 20. The number of mice in a group is acceptable.

 

Comments 3: The "2" in "H2O" in Figure 1A should be marked as a subscript to conform to the standard expression of the chemical formula, and the same is true for the caption.

Response 3: Figure 1A has been corrected.

 

Comments 4: Why is the resolution of Figure 2 E too low?

Response 4: Figure 2E has been replaced with a similar high-resolution version.

 

Comments 5: It was observed that the 81 mmol/L group in Figure 1A showed 7 data points, while the other groups showed 8. Please check the data and state whether there are any missing data or plotting errors.

Response 5: There was one outlier value in this group (81 mmol/L group), which was excluded.

 

Comments 6: The color scheme, axis format, error bar representation, and other aspects of the charts and graphs in the full text are inconsistent in style, which affects the visual professionalism and coherence. It is recommended to unify the color scheme (such as using the color palette recommended by the journal) and standardize the format elements of all charts and graphs.

Response 6: All figures representing BLI data were unified and re-uploaded.

 

Comments 7: The authors explicitly did not use microbubbles, citing "increased cost and complexity." However, the vast majority of literature on efficient sonoporation relies on microbubbles to enhance cavitation. This makes the study an evaluation of a "non-standard" or "simplified" version of the sonoporation effect. The authors should add the key limiting condition of "no microbubbles" in the abstract and conclusion to avoid misleading readers' overall understanding of ultrasound perforation technology.

Response 7: Indeed, we have not used microbubbles in our study. Our study aimed to find a technically simple method for DNA delivery/DNA immunization using available inexpensive clinically used equipment, not relying on the use of additional reagents which would generate additional costs.

Sonoporation refers to temporal enhancement of the permeability of cellular membranes due to cavitation caused by the ultrasound and acoustic waves (https://pubmed.ncbi.nlm.nih.gov/27094209/ ). Cavitation is a nonthermal interaction between a propagating pressure wave and gaseous inclusions (microscopic bubbles or cavities) in a liquid (https://pubmed.ncbi.nlm.nih.gov/12774945/ ). Sub-micrometer gas bodies (“microbubbles”) exist naturally in blood and tissue and behave as endogenous cavitation nuclei (https://pubmed.ncbi.nlm.nih.gov/17042253/ , https://pubmed.ncbi.nlm.nih.gov/12774945/ ). Low-pressure acoustic amplitudes cause stable pulsation of gas bubbles (https://pubmed.ncbi.nlm.nih.gov/12774945/ ). When the acoustic pressure amplitudes are increased, a violent collapse called inertial cavitation occurs, accompanied by an increase in the size of the natural microbubbles, followed by their explosion and emergence of  shock waves and microstreaming, which altogether lead to the formation of temporary holes in the cell membrane (https://www.mdpi.com/1999-4923/14/8/1642 ). By varying the parameters of ultrasound parameters, one can change the cavitation effects (size and behavior of microbubbles) and, consequently, the overall impact on the cell membrane. This text and links are added to the INTRODUCTION section to describe the process of sonoporation (page 3, inserted with tracking, shaded gray). Not using “exogenous” microbubbles does not simplify the process, as the Reviewer suggests. It still depends on the formation and fate of microbubbles.

The use of additional (external) microbubbles makes the delivery by sonoporation independent of the natural process of microbubble formation. With the use of “exogenous” microbubbles, the process would mostly rely on the capacity of sonic waves to burst the existing/added bubbles and make holes in cell membranes. Addition of the "exogenous" microbubbles increases the impact of sonoporation on the cell membrane (pore formation) without the necessity to increase the intensity of acoustic waves, otherwise needed to generate more “natural” microbubbles. However, a balance must be maintained, as exceeding the threshold level of microjet exposure from the additional “exogenous” microbubbles will cause irreversible damage of cell membranes and cell death by lysis. I.e., application of "exogenous" microbubbles requests optimization of both the dose and the size of added microbubbles. Both parameters vary depending on the origins/nature of the tissues to be targeted by sonication (see for example, https://www.mdpi.com/2227-9059/9/7/803 , https://www.sciencedirect.com/science/article/pii/S1350417719318413#b0060 ). There are also issues of histocompatibility and toxicity of the  “exogenous” microbubbles, as well as the potential harm caused by the long-term accumulation of the breakdown products of “exogenous” microbubbles in the body (https://www.mdpi.com/1999-4923/14/8/1642).

In our view, dependence on the provision of microbubbles, related extra costs, histocompatibility of microbubbles as well as toxicity limit the use of sonoporation with “exogenous” microbubbles.  We attempted to overcome these limitations (remove extra variables and extra costs) by optimizing sonoporation protocol.

These considerations motivating our efforts to optimize sonoporation as such, free of the use of “exogenous” microbubbles, were included in the INTRODUCTION section (pages 3, inserted with tracking, shaded gray) and DISCUSSION section (pages 24, inserted with tracking, shaded gray).

 

Comments 8: The control of variables is not rigorous enough: there are differences in DNA dosage (50μg, 100μg) and the number of injection sites (1, 2) in the experiment. Although the standardization method is used to try to correct it, it may still affect the direct comparability between different delivery methods, and the rationality of the variable setting is not clearly stated.

Response 8: Authors wish to thank the Reviewer for raising the sensitive issue of varying numbers of injections, and varying DNA dosage.

The first ID series was performed with two DNA injections per mouse, with DNA dose of 100 mcg per injection. As noted on study limitations, we had used high doses of DNA, up-to 100 µg per injection site, to maximize the DNA-modulated decrease of the ionic strength of delivery buffer, to be able to detect and characterize the effect. In ID series, DNA dose was increased to maximal which can be still delivered intradermally (volume limitation) as, based on the earlier data by us and by other groups, we expected low expression levels of Luc after not assisted/not facilitated DNA delivery. We faced difficulties in delivering this high dose due to high viscosity of the solution. Delivery of such solution through thin needle as in insulin syringe can lead to DNA fragmentation, as was noted in the limitations of the study. Hence, in all subsequent experiments with facilitated DNA delivery (where we expected stronger bioluminescent signals due to higher levels of Luc expression after sonoporation/SP or electroporation/EP) the dose of DNA was decreased from 100 to 50 mcg. In case of EP, this approach worked well, as bioluminescence signal generated after delivery of 50 mcg of Luc DNA was strong, and could be reliably detected. In case of SP, the result was not as good as we expected, SP did not lead to an improvement of Luc gene delivery and subsequent increase Luc expression, BLI signals were relatively low. The result could have been better if 100 mcg were used in the repeated experiment. However, a decision was made not to repeat the experiment with higher DNA doses, as overall, SP was not shown to improve Luc expression after ID delivery. A repeat of the failing experiments would be against the 3R rules for animal research.

Regarding the number of injections, in case of delivery of Luc gene, increasing number of injection sites allowed to increase the number of observations without increasing the number of animals, which answered the general 3R requirements for animal research. We exploited this option in delivery of Luc gene in ID series (5 mice per group generating 10 observations).

All other experiments included a physical treatment of the animal – sonoporation, or electroporation applied on top of ID injection. Both treatments affect the animal, causing local inflammation, such as swelling for SP and local burns for EP, which could potentially spread to tissues neighbouring the site of treatment (SP or EP) including the site of second injection (in a mouse body, two injections sites are separated by max 2-2,5 cm). Local inflammation would interfere with gene expression which relies on the minimal interference of the innate immune response. Hence, all experiments with facilitated ID injections were performed at one site only. An exclusion was one experiment in EP series (EP-2), where DNA was injected into two sites with subsequent electroporation. This was done to compare the results of ID delivery alone (done at two sites) with ID delivery reinforced by electroporation.

We have added a detailed motivation for using varying numbers of injections, and varying DNA dosage to the LIMITATIONS section (pages 25-26, inserted with tracking, shaded gray).

Reviewer 2 Report

Comments and Suggestions for Authors

In this manuscript, the authors systematically investigate how different administration techniques affect the performance of DNA vaccines. The study is logically structured, the results are well-presented, and the findings hold potential clinical relevance. I only have a few minor comments:

1. In some experimental groups, the authors used different numbers of injection sites. Would this variation influence the outcomes? The authors should discuss this point.

2. The authors should include an appropriate blank control group, such as animals receiving the injection procedure without active DNA, to demonstrate any responses induced solely by the injection itself.

Author Response

Comments 1: In some experimental groups, the authors used different numbers of injection sites. Would this variation influence the outcomes? The authors should discuss this point.

Response 1: The number of injections is indeed a very important aspect of our experiments, questioned by both Reviewer #1 and Reviewer #2. Extended answer was given in response to the critical issue 8 raised by the Reviewer #1. It is repeated here for convenience.

In case of delivery of Luc gene, increasing number of injection sites allowed to increase the number of observations without increasing the number of animals, which answered the general 3R requirements for animal research. We exploited this option in delivery of Luc gene in ID series (5 mice per group generating 10 observations).

All other experiments included a physical treatment of the animal – sonoporation, or electroporation applied on top of ID injection. Both treatments could affect the animal, causing local inflammation, such as swelling for SP and local burns for EP, which could potentially affect tissues neighbouring the treatment site, including the site of second injection (in a mouse body, two injections sites are separated by max 2-2,5 cm). Local inflammation could have affected the results of experiments aiming the assess the levels of gene expression with minimal interference of innate immune response. In view of these, all experiments with reinforced ID injections were performed at one site only. One exclusion was experiment EP-2 in EP series, where DNA was injected into two sites with subsequent electroporation. This was done to compare the results of ID delivery alone with ID delivery reinforced by electroporation. These considerations were included in the LIMITATIONS section (page 26, inserted with tracking, shaded gray).

 

Comments 2: The authors should include an appropriate blank control group, such as animals receiving the injection procedure without active DNA, to demonstrate any responses induced solely by the injection itself.

Response 2: Our study had two arms, one aimed at optimization of delivery by intradermal injections (ID), or ID reinforced by sonoporation (SP) or electroporation (EP). In the respective series we varied conditions for delivery, choosing those ensuring the best delivery, selected based on the strength of bioluminescence due to expression of Luc gene. Measurements in these series were relative, and as such did not require a blank.

Furthermore, none of the treatments (ID, SP, EP) was expected to produce any bioluminescence. The only background expected in these experiments is autoluminescence, a natural background light emitted by endogenous fluorophores within living tissues (keratin, collagen, elastin, NAD(P)H, porphyrins) excited by an external light source. It is a persistent background signal, induced by neither mechanical actions such as injections, nor by injections of non-luminescent substances.

Based on these considerations, blank control groups showing bioluminescence in untreated groups (groups not receiving Luc gene, but receiving blank ID injections, or ID/SP, or ID/EP treatments) provided no additional information, and were therefore, not included.

At the same time, we realized that it is important to demonstrate if registered signals are specific, i.e. are above the background. For this, in all series, we included data from a blank control group of naive mice not treated with either ID, or ID/SP, or ID/EP, but injected with D-luciferin showing the level of autoluminescence. This background is shown in Figure 1 for ID series, in Figure 2 in SP series, and in Figure 3 in EP series, respective text in the figure legends is shaded yellow.

The second arm of the study aimed at applying the best delivery conditions in DNA immunization, to induce immune response against viral protein encoded by a plasmid, here reverse transcriptase of HIV-1 (RT). The readout was the level of anti-RT antibody response. In these experiments, it was important to have a blank control group to have a read-out of the background signal - unspecific reactivity of murine sera against recombinant RT. Our earlier studies on the induction of immune response against viral proteins by DNA immunization demonstrated similar behaviour of the control groups receiving no treatment (naïve mice), or receiving deliveries of water, or non-coding DNA (empty vector). Hence, series IR was shaped to include a control group of untreated (naïve) mice.