The Role of NMDAR and BDNF in Cognitive Dysfunction Induced by Different Microwave Radiation Conditions in Rats
, Hongmei Zhou 1, Ruiyun Peng 1, Xiangjun Hu 1, Yong Zou 1,* and Lifeng Wang 1,4,*Round 1
Reviewer 1 Report
The authors investigated the relationship between changes of brain function by microwave radiation and its mechanisms. The topic addressed are important for understanding mechanisms of normal tissue damage by radiation.
Although I have no doubt about the quality of the presented work, I recommend to revise the manuscript about some points.
- Figure 1A: How about the data of 7d and 14d? The functional changes of brain are induced by changes of neurotransmitters. This manuscript showed that the several neurotransmitters changed 7d after radiation. I think that the authors have to do the MWM test until 7d or 14d.
- Figure 2: Please indicate focus points in this figure.
- Figure 3: Why are the expression level of 0 mW/cm2 (i.e. control) difference in 6h, 3d and 7d?
- Figure 3 and 4: The expression of neurotransmitters and proteins were time and dose independent. Please consider that.
- Page 9, line 14: Please correct it. “…microwave-associated synaptic plasticity impairment24.” → “…microwave-associated synaptic plasticity impairment [24].”
Author Response
A POINT BY POINT REPLY TO REVIEWER 1
Comments and Suggestions for Authors
The authors investigated the relationship between changes of brain function by microwave radiation and its mechanisms. The topic addressed are important for understanding mechanisms of normal tissue damage by radiation.
Although I have no doubt about the quality of the presented work, I recommend to revise the manuscript about some points.
1. Figure 1A: How about the data of 7d and 14d? The functional changes of brain are induced by changes of neurotransmitters. This manuscript showed that the several neurotransmitters changed 7d after radiation. I think that the authors have to do the MWM test until 7d or 14d.
Response: In the MWM test, the navigation and probe trials were used. The navigation abilities were tested in 6h-3d, and probe trials were carried out at 14 d after microwave radiation. The probe trials only used one time to avoid memory confusion in rats. So we detected the probe trials at 14d following the results of neurotransmitters and TEM.
2. Figure 2: Please indicate focus points in this figure.
Response: We have added arrows showing the locations of PSDs in Figure 2.
3. Figure 3: Why are the expression level of 0 mW/cm2 (i.e. control) difference in 6h, 3d and 7d?
Response: Because 0 mW/cm2 group was not the blank control group, the rats in the 0 mW/cm2 group were placed in the irradiation box on a radiation table but were not exposed to microwave radiation. The different expression levels of 0 mW/cm2 in 6h, 3d and 7d might show the effects of the operation during irradiation and environmental effects on the contents of amino acids neurotransmitters.
4. Figure 3 and 4: The expression of neurotransmitters and proteins were time and dose independent. Please consider that.
Response: In the experiments, we detected the expression of amino acids neurotransmitters including Asp, Glu, GABA and Gly which might were related to learning and memory ability and hippocampal structure. NMDAR and the associated proteins were activated by glutamic acids. In the Figure 4, we detected the expression of NMDAR, which showed the increased expression in 30 mW/cm2 group which was consistent with the change of glutamate, and the decreased expression in 50 mW/cm2 group. The time and dose independence might be related to the participation of other molecules.
5. Page 9, line 14: Please correct it. “…microwave-associated synaptic plasticity impairment24.” → “…microwave-associated synaptic plasticity impairment [24].”
Response: Thanks for your suggestion. We have corrected it in the manuscript.
Author Response File: 
Author Response.docx
Reviewer 2 Report
This manuscript reports experiments on young adult rats exposed to pulsed microwaves and analyzed from 6 hr to 14 days later for effects on the central nervous system, specifically the hippocampus. The study design is similar to prior studies and is confirmatory rather that breaking new ground. There are a few issues with the behavioral methods that are discussed below but biochemical and microscopy methods are of high quality. A few refinements to the discussion would improve the manuscript which is acceptable for publication with minor revisions and clarifications.
Microwaves
One question that arises is why the frequency of 2.856 GHz was chosen. This frequency is associated with clinical linear accelerators but is generally higher than frequencies used by mobile communications devices and other commercial products. So, if the intent is to estimate health risks, versus using a convenient S-band model giving resonant absorption, some short explanation of this selection would clarify intent. The literature indicates that frequencies from 1 to 6 GHz are behaviorally effective. The methods indicate that the microwaves were pulsed but there is no detail on the modulation. The pattern of modulation has been shown to be important with the possibility of interacting with rhythmic brain activity at low frequencies and on short time scales may drive protein denaturation or ion movements (for example papers by D’Andrea) . Please add more detail. Also, it would be helpful to indicate whether the authors regard the study as being affected by thermal or non-thermal effects. Some authors argue that behavioral responses may be secondary to animals simply reacting to surface or core temperature changes.
Morris Water Maze
The protocol for the water maze is non-standard and lacks some detail. While there are curtains in the test room it isn’t clear whether there are external visual cues. What is the overall lighting level and sound environment? Is the water made opaque with paint, powdered milk, etc.? There is no cued trial, entry points are not randomized and the probe trial is delayed for quite a while relative to the last training trial. The sequential starting points by quadrant could provide information (rules, like “always turn to the right”) to the animal for identifying platform location that is not strictly spatial and therefore may involve non-hippocampal brain regions. External cues would also affect the degree of spatial learning and memory. The definition of “crossings” for figure 1B needs to be clarified. Does “crossing” mean crossings of the platform zone only or of the quadrant that the hidden platform occupies. The pooling of data for figure 1 A needs more explanation and the variance should be displayed (error bars). Do the data points represent the average of each of the four trials for every animal? This could be problematic, as when animals are placed in the first quadrant, where the platform is located, they would likely have a much lower escape latency than from other more distant entry sites – this could distort the data. Similarly, if the entry points on the four trial series are always in the same order, there could be systematic effects on latency. Please clarify the averaging strategy and update the figure with error bars. Lastly, what time of day were the behavioral tests performed? This should be consistent with respect to light:dark cycle. For a 6 hour post irradiation time point this would seem likely to be an afternoon time whereas the 1, 3, 7, 14 day points are most likely morning times. Is there evidence that the 6 h time point is influenced by circadian rhythms different from the other time points.
Electron Microscopy
The interpretation of EM micrographs requires some practice. On figure 2 please add some arrows showing the locations of interest in the photos. In describing the width and thickness of PSDs is this a surrogate for the frequency of mushroom versus short dendritic spines or filopodia? With ionizing radiation, filopodia (small PSDs) numbers decrease leading to a shift to a higher proportion of mushroom spines (larger PSDs). Is it possible to determine whether the spine numbers are constant but changing in morphology or whether the distribution of spine types has changed? It would be expected that decreased spine numbers and dendritic complexity would lead to impaired performance. PSD density itself would be associated with altered plasticity. Is it possible that microwaves elicit inappropriate synaptic pruning by microglial activation? The presence of denatured proteins from thermal effects might lead to such responses.
Neurotransmitters
The neurotransmitter response is not easy to interpret. Asp and Glu excitatory transmitter and Gly and GABA inhibitory transmitter responses seem inconsistent. Opposite time course responses for the members of each pair are observed. Are these levels related to presynaptic release that may have been driven by microwave-induced ion movements (e.g. calcium) or do they reflect overall metabolism or astrocyte dysfunction in maintaining glutamate levels in the synaptic cleft?
NMDAR, Cortactin and BDNF/TrkB
The Western blots look good. lines 319 – 327 emphasize PSD95/cortactin and NMDAR expression levels but their topology is also important even at constant concentration. Other related features relevant to synaptic plasticity are recruitment and turnover of AMPA receptors to the postsynaptic region and the phosphorylation status of NMDAR’s.
While BDNF activates NMDARs it isn’t clear that over-activation of BDNF causes cytotoxicity (line 331). Isn’t excitotoxicity due to excess glutamate build-up then causing excessive cell depolarization accompanied by oxidative stress, etc.
Discussion
The comparisons in lines 247 – 263 are for studies with distinctly different microwave frequencies, power densities etc. and times post irradiation so the relationships of these findings to the present study must be interpreted with caution. On the other hand, do the studies indicate collectively that behavioral effects are generally S-band microwave-sensitive and that the details of the exposure are of secondary concern?
line 273 – ref 24 should be in brackets “[24]”.
line 276 – “was significantly at 6 months” is incomplete
lines 289 – 291. The authors have correctly and honestly pointed out the inconsistencies between the current and prior studies with respect to neurotransmitter levels. Thank-you for the candor. However, it doesn’t seem particularly likely that the difference in exposure time from 5 to 6 minutes with other parameters being the same would drive the differences – 15 – 20% cumulative dose difference. Were both studies using the same pulse modulation as well as the power density?
Author Response
A POINT BY POINT REPLY TO REVIEWER 2
Comments and Suggestions for Authors
This manuscript reports experiments on young adult rats exposed to pulsed microwaves and analyzed from 6 hr to 14 days later for effects on the central nervous system, specifically the hippocampus. The study design is similar to prior studies and is confirmatory rather that breaking new ground. There are a few issues with the behavioral methods that are discussed below but biochemical and microscopy methods are of high quality. A few refinements to the discussion would improve the manuscript which is acceptable for publication with minor revisions and clarifications.
Microwaves
One question that arises is why the frequency of 2.856 GHz was chosen. This frequency is associated with clinical linear accelerators but is generally higher than frequencies used by mobile communications devices and other commercial products. So, if the intent is to estimate health risks, versus using a convenient S-band model giving resonant absorption, some short explanation of this selection would clarify intent. The literature indicates that frequencies from 1 to 6 GHz are behaviorally effective. The methods indicate that the microwaves were pulsed but there is no detail on the modulation. The pattern of modulation has been shown to be important with the possibility of interacting with rhythmic brain activity at low frequencies and on short time scales may drive protein denaturation or ion movements (for example papers by D’Andrea). Please add more detail. Also, it would be helpful to indicate whether the authors regard the study as being affected by thermal or non-thermal effects. Some authors argue that behavioral responses may be secondary to animals simply reacting to surface or core temperature changes.
Response: The details of the microwave exposure system were elaborated in a previous report [23,24], so in the manuscript, we don’t describe it in detail. The pattern of modulation has been added in the manuscript, in which pulse width was 1μs and repetition frequency was 250Hz. We have detected the anal temperature during the exposure, but the data have not been published. For 10 mW/cm2 group, there was no temperature rise (0.04°C); for 30 mW/cm2 group, there was a temperature rise of 1.03 °C; for 50 mW/cm2 group, there was a temperature rise of 1.35 °C.
Morris Water Maze
The protocol for the water maze is non-standard and lacks some detail. While there are curtains in the test room it isn’t clear whether there are external visual cues. What is the overall lighting level and sound environment? Is the water made opaque with paint, powdered milk, etc.? There is no cued trial, entry points are not randomized and the probe trial is delayed for quite a while relative to the last training trial. The sequential starting points by quadrant could provide information (rules, like “always turn to the right”) to the animal for identifying platform location that is not strictly spatial and therefore may involve non-hippocampal brain regions. External cues would also affect the degree of spatial learning and memory. The definition of “crossings” for figure 1B needs to be clarified. Does “crossing” mean crossings of the platform zone only or of the quadrant that the hidden platform occupies. The pooling of data for figure 1 A needs more explanation and the variance should be displayed (error bars). Do the data points represent the average of each of the four trials for every animal? This could be problematic, as when animals are placed in the first quadrant, where the platform is located, they would likely have a much lower escape latency than from other more distant entry sites – this could distort the data. Similarly, if the entry points on the four trial series are always in the same order, there could be systematic effects on latency. Please clarify the averaging strategy and update the figure with error bars. Lastly, what time of day were the behavioral tests performed? This should be consistent with respect to light:dark cycle. For a 6 hour post irradiation time point this would seem likely to be an afternoon time whereas the 1, 3, 7, 14 day points are most likely morning times. Is there evidence that the 6 h time point is influenced by circadian rhythms different from the other time points.
Response: There were some visual cues on the walls of Morris Water Maze, such as â–³, ┿, etc. It was lighting and quiet during the test, and the water was without any other supplements. When the rats entered water, they faced the maze wall. “Crossing” means crossings of the platform zone. The data points represent the average of each of the four trials for every animal. And we mainly focused on the difference of radiation groups and 0 mW/cm2 group. The probe trials only used one time to avoid memory confusion in rats. So we detected the probe trials at 14d following the results of neurotransmitters and TEM. For the point of 6h, it should be 0h. The experiment started immediately after irradiation, but because it was the first test and took a long time for about 8 hours. And we have corrected the errors and added the necessary information in the manuscript.
Electron Microscopy
The interpretation of EM micrographs requires some practice. On figure 2 please add some arrows showing the locations of interest in the photos. In describing the width and thickness of PSDs is this a surrogate for the frequency of mushroom versus short dendritic spines or filopodia? With ionizing radiation, filopodia (small PSDs) numbers decrease leading to a shift to a higher proportion of mushroom spines (larger PSDs). Is it possible to determine whether the spine numbers are constant but changing in morphology or whether the distribution of spine types has changed? It would be expected that decreased spine numbers and dendritic complexity would lead to impaired performance. PSD density itself would be associated with altered plasticity. Is it possible that microwaves elicit inappropriate synaptic pruning by microglial activation? The presence of denatured proteins from thermal effects might lead to such responses.
Response: We have added arrows showing the locations of PSDs in Figure 2. We have detected the effects of microwave on the dendritic spines and the mechanism, which has been published as follows: Microwave Radiation Leading to Shrinkage of Dendritic Spines in Hippocampal Neurons Mediated by SNK-SPAR Pathway. Brain Research, 2018, 1679:134-143.
Neurotransmitters
The neurotransmitter response is not easy to interpret. Asp and Glu excitatory transmitter and Gly and GABA inhibitory transmitter responses seem inconsistent. Opposite time course responses for the members of each pair are observed. Are these levels related to presynaptic release that may have been driven by microwave-induced ion movements (e.g. calcium) or do they reflect overall metabolism or astrocyte dysfunction in maintaining glutamate levels in the synaptic cleft?
Response: Thanks for your suggestion. We have detected the release of presynaptic synaptic vesicles and the mechanism, which has been published as follows:
- Reduction of Phosphorylated Synapsin I (Ser-553) Leads to Spatial Memory Impairment by Attenuating GABA Release after Microwave Exposure in Wistar Rats. PLOS one, 2014; 9(4): e95503
- Abnormality of synaptic vesicular associated proteins in cerebral cortex and hippocampus after microwave radiation. Synapse, 2009; 63(11): 1010-1016.
NMDAR, Cortactin and BDNF/TrkB
The Western blots look good. lines 319 – 327 emphasize PSD95/cortactin and NMDAR expression levels but their topology is also important even at constant concentration. Other related features relevant to synaptic plasticity are recruitment and turnover of AMPA receptors to the postsynaptic region and the phosphorylation status of NMDAR’s.
While BDNF activates NMDARs it isn’t clear that over-activation of BDNF causes cytotoxicity (line 331). Isn’t excitotoxicity due to excess glutamate build-up then causing excessive cell depolarization accompanied by oxidative stress, etc.
Response: Thanks for your suggestion. We will focus on the topology of PSD95/cortactin and NMDAR in the following study. It showed that over-activation of NMDAR causes cytotoxicity, but BDNF activates NMDARs in which BDNF mainly play a protective role.
Discussion
The comparisons in lines 247 – 263 are for studies with distinctly different microwave frequencies, power densities etc. and times post irradiation so the relationships of these findings to the present study must be interpreted with caution. On the other hand, do the studies indicate collectively that behavioral effects are generally S-band microwave-sensitive and that the details of the exposure are of secondary concern?
Response: Studies have shown that exposure to excessive microwave radiation can cause dysfunction of learning and memory ability. Behavioral effects are sensitive to the microwave radiation, but the exposure parameters of microwave are more important and the effects occur when the exposure dose exceeds the safety limit (SAR 4W/Kg).
line 273 – ref 24 should be in brackets “[24]”.
Response: We have corrected it in the manuscript.
line 276 – “was significantly at 6 months” is incomplete
Response: We have corrected it in the manuscript as “was significantly decreased at 6 months”.
lines 289 – 291. The authors have correctly and honestly pointed out the inconsistencies between the current and prior studies with respect to neurotransmitter levels. Thank-you for the candor. However, it doesn’t seem particularly likely that the difference in exposure time from 5 to 6 minutes with other parameters being the same would drive the differences – 15 – 20% cumulative dose difference. Were both studies using the same pulse modulation as well as the power density?
Response: Although the average power density of the two experiments is the same, the pulse modulations are different which may be the main reason for the different effects, not the difference of irradiation time for 5min or 6min.
Author Response File: Author Response.docx
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
