Astrocyte-Conditioned Medium Induces Protection Against Ischaemic Injury in Primary Rat Neurons

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In this study, the authors investigate the time-dependent responses of astrocytes to OGD and whether astrocyte conditioned medium can protect primary rat neurons when applied prior to OGD. Unsurprisingly they discovered that shorter durations of OGD upregulated putative protective genes without a loss of cells, while prolonged OGD led to a stress response. While it is interesting that different OGD times produce ACM with different effects on neurons enthusiasm for the study is dampened by the fact that the observation would have benefited from a more comprehensive analysis. For example, a better characterization of transcriptional changes using RNAseq (this would at least show important pathways rather than single genes, which are hard to infer anything from without functional knockouts), and better cell characterization. It is already known that the astrocyte response is contextual and time-dependent.

Line 16: I didn’t think Epo was upregulated in 6h OGD as shown in Figure 5.

Line 283: Misspelt “microscopic filed”?

Figure 2A: The images are very dim. It’s difficult to see the GFAP and the DAPI is not evident at all? Whether the GFAP colocalizes with DAPI is unknown.

Figure 2B: Is the scale on the y-axis correct? So the GFAP intensity of OGD 24h went up by only approx 5% of control (Nx). In Panel A the image looks like it went up by far far more, a huge difference (e.g., 24 h Nx versus 24 h OGD).

Figure 2C: Is it just a loss of GFAP immunoreactivity or are the cells dead?

Figure 4A: What is OD?

Figure 5: It is worth mentioning earlier in the manuscript why the authors have chosen the genes they did. It is in the discussion, but that seems too late.

Line 352, 353, 356, 357: I presume the “n” is number animals/litters, but within the culture, how many neurons were examined?

Line 354: “preserved” should probably be “prevented”

Section 3.2: Could you measure some other characteristic of neurons other than simply dendrite length e.g., oxidative stress? Neuroprotective potential shouldn’t just be based on dendrites per neuron. Also, do you see an increase in the number of neurons survive with the preconditioning using 6h OGD-ACM?

Figure 6: Again here the images are low quality and look blurry. Also, the DAPI is not visible in most of the panels.

Can you speculate what it is about the 24 hr OGD ACM that does not confer neuroprotection to the OGD exposed primary rat neurons, compared to the 6hr OGD ACM?

Author Response

In this study, the authors investigate the time-dependent responses of astrocytes to OGD and whether astrocyte conditioned medium can protect primary rat neurons when applied prior to OGD. Unsurprisingly they discovered that shorter durations of OGD upregulated putative protective genes without a loss of cells, while prolonged OGD led to a stress response. While it is interesting that different OGD times produce ACM with different effects on neurons enthusiasm for the study is dampened by the fact that the observation would have benefited from a more comprehensive analysis. For example, a better characterization of transcriptional changes using RNAseq (this would at least show important pathways rather than single genes, which are hard to infer anything from without functional knockouts), and better cell characterization. It is already known that the astrocyte response is contextual and time-dependent.

Thank you for your thoughtful and insightful feedback on our manuscript.

We agree to your comments and add a new paragraph in the section of “Limitation and Future Directions” as follows:

The finding that different durations of OGD yield ACM with distinct effects is intriguing and suggests a dynamic, context-dependent astrocyte response. However, the overall strength of the conclusions could be enhanced with more comprehensive analysis. For instance, RNA sequencing could have provided a broader and more informative overview of transcriptional changes, revealing key pathways rather than focusing on a few selected genes, which offer limited insight without functional validation. Additionally, more thorough characterization of the astrocyte populations would strengthen the findings. Given that the astrocyte response is already well-documented as being both contextual and time-dependent, a deeper mechanistic analysis is essential to substantiate the study's contribution to the field.

Line 16: I didn’t think Epo was upregulated in 6h OGD as shown in Figure 5.

Thanks for spotting this error. We now remove Epo from this sentence in Line 16.

Line 283: Misspelt “microscopic filed”?

Well spotted. We have corrected this typo to “microscopic field”.

Figure 2A: The images are very dim. It’s difficult to see the GFAP and the DAPI is not evident at all? Whether the GFAP colocalizes with DAPI is unknown.

We thank the reviewer for their observation regarding the immunofluorescence image quality. We acknowledge that the current images appear dim, particularly in terms of GFAP and DAPI signal intensity, which may make it difficult to assess nuclear staining and potential colocalization.

To address this concern, we have reprocessed the images with optimized contrast and brightness settings using standardized image processing protocols (without altering the raw data) to enhance the visibility of both GFAP and DAPI channels. These revised images have been included in the updated manuscript (Figure 2).

We also confirm that GFAP-positive cells exhibit DAPI nuclear staining, consistent with astrocyte identity. While GFAP is a cytoplasmic marker and does not colocalize with DAPI per se, the merged images demonstrate that GFAP signal surrounds DAPI-positive nuclei, indicating correct labeling of astrocytes. We have clarified this point in the figure legend.

We appreciate the reviewer’s helpful feedback and believe the revised figure presentation improves interpretability and clarity.

Figure 2B: Is the scale on the y-axis correct? So the GFAP intensity of OGD 24h went up by only approx 5% of control (Nx). In Panel A the image looks like it went up by far far more, a huge difference (e.g., 24 h Nx versus 24 h OGD).

Thanks for spotting this wrong label. It should be “folds to control”. We have changed.

Figure 2C: Is it just a loss of GFAP immunoreactivity or are the cells dead?

We have seen a loss of GFAP immunoreactivity, but did not see the loss of nuclear DAPI staining. Prolonged OGD can suppress GFAP expression in astrocytes

Figure 4A: What is OD?

OD means “oxygen deprivation”. We have added the full name in the Figure4 legend. Thanks.

Figure 5: It is worth mentioning earlier in the manuscript why the authors have chosen the genes they did. It is in the discussion, but that seems too late.

We have mentioned why we had chosen the genes in the methods 2.8 as follows

Actin was used as an internal control to normolise the relative levels of mRNA, while others were responsive to hypoxia. Their sequences were listed in Table 1.

Line 352, 353, 356, 357: I presume the “n” is number animals/litters, but within the culture, how many neurons were examined?

Yes, the “n” is number animals/litters. We examined several neurons per microscopic field.

Line 354: “preserved” should probably be “prevented”

Yes. We have made changes. Thanks.

Section 3.2: Could you measure some other characteristic of neurons other than simply dendrite length e.g., oxidative stress? Neuroprotective potential shouldn’t just be based on dendrites per neuron. Also, do you see an increase in the number of neurons survive with the preconditioning using 6h OGD-ACM?

We thank the reviewer for this valuable and thoughtful suggestion. We agree that dendritic length alone may not fully capture the extent of neuroprotection and that incorporating additional markers of neuronal viability and function would provide a more comprehensive assessment. In this study, our primary aim was to assess morphological preservation as a preliminary indicator of neuroprotection following exposure to astrocyte-conditioned medium (ACM). Dendritic length was chosen due to its sensitivity to stress and injury in primary neurons. However, we fully acknowledge the limitations of using this single metric and appreciate the reviewer’s recommendation to include complementary assays such as oxidative stress markers (e.g., ROS levels), mitochondrial function (e.g., JC-1, ATP assays), or cell death pathways (e.g., caspase-3 activity, TUNEL staining).

Figure 6: Again here the images are low quality and look blurry. Also, the DAPI is not visible in most of the panels.

We thank the reviewer for pointing out the issue regarding image quality and DAPI visibility. We acknowledge that in the original submission, the resolution and contrast of the immunofluorescence panels may have been insufficient to clearly visualize the DAPI signal in certain images.

To address this, we have reprocessed the images using standardized adjustments (e.g., contrast enhancement and background subtraction) to improve clarity while ensuring no manipulation of the raw signal. The DAPI signal is now clearly visible in all conditions and better demonstrates the nuclear localization relative to the GFAP-positive astrocytic processes. We have updated the figure in the revised manuscript (Figure 6) and clarified this in the figure legend.

We appreciate the reviewer’s attention to detail and believe the revised figures significantly improve visual clarity and interpretability.

Can you speculate what it is about the 24 hr OGD ACM that does not confer neuroprotection to the OGD exposed primary rat neurons, compared to the 6hr OGD ACM?

We thank the reviewer for their insightful comment regarding the differential effects of 6-hour versus 24-hour OGD AC). We agree that the underlying mechanisms are complex and warrant further investigation. As the reviewer correctly notes, astrocyte responses to injury are highly time-dependent. Based on current literature and our observations, we speculate that shorter OGD durations induce a preconditioning-like state in astrocytes, promoting the release of neuroprotective factors such as VEGF, IL-10, and exosomes enriched with supportive cargo [Diaz-Guerra et al., 2020;  Lee et al., 2022; Zhang et al., 2024]. In contrast, prolonged OGD may lead to cellular stress or dysfunction, potentially shifting the secretome toward pro-inflammatory cytokines (e.g., TNF-α, IL-6), glutamate, and metabolic byproducts that could exacerbate neuronal injury [Liddelow et al., 2017]. Additionally, astrocytic metabolic support may become compromised under extended hypoxia, reducing their ability to supply protective intermediates such as lactate [Jha and Morrison, 2020]. Maladaptive autophagy or senescence-like phenotypes could further contribute to this loss of neuroprotective capacity [Wu et al., 2021]. We acknowledge the limitations in our current dataset and plan to address these mechanisms in future studies using transcriptomic profiling (e.g., RNA-seq) and functional assays to better delineate the molecular drivers of ACM-mediated neuroprotection.

Lee H, Yoon Y, Lee M. Neuroprotective role of astrocyte-derived exosomes in oxygen-glucose deprivation-induced neuronal injury. Biochem Biophys Res Commun. 2022;606:110–117. doi:10.1016/j.bbrc.2022.04.105

Díaz-Guerra M, Velasco A, Avila J, Hernández F. Inflammatory response in preconditioning: A new approach to a classical mechanism. Int J Mol Sci. 2020;21(2):556. doi:10.3390/ijms21020556

Zhang W, Liang C, Guo S, Huang Y, Liang G. Therapeutic potential of astrocyte-derived extracellular vesicles in ischemic stroke: Mechanisms and future prospects. Aging Dis. 2024. doi:10.14336/AD.2023.0823-1

Liddelow SA, Guttenplan KA, Clarke LE, et al. Neurotoxic reactive astrocytes are induced by activated microglia. Nature. 2017;541(7638):481–487. doi:10.1038/nature21029

Jha MK, Morrison BM. Lactate transporters mediate glia-neuron metabolic crosstalk in homeostasis and disease. Front Cell Neurosci. 2020;14:589582. doi:10.3389/fncel.2020.589582

Wu Q, Zhang X, Shao M, Zhang C, Sun Y, Liu J, Wang Q, Huang Y, Wang T. Dysfunctional autophagy in astrocytes contributes to chronic stress-induced cognitive impairments. Cell Death Dis. 2021;12(8):760.

Reviewer 2 Report

Comments and Suggestions for Authors

Question.

1. Line 147. How did you control that the astrocytes reached confluency?

Comments.

2. The references list and in-text refs are formatted not according to the journal rules. See Reference List and Citations Style Guide for MDPI Journals on the journal website.
3. Figure 4B does not contain column names.
4. Line 390, 401. Authors claim that the Lc3b-II expression was significantly lower after 24 h OGD-ACM treatment than that after 6 h OGD-ACM treatment. However, the confidence intervals of the obtained data overlap, that does not allow make this statement for sure. The same for Glut1 and Pfkfb3 upregulation on figure 5.
5. Despite on unconfident Lc3b-II upregulation, p62 protein still downregulated, that witnessed autophagy occurs in OGD astrocytes. Based on that, authors suggest autophagy as a key mechanism in ACM-mediated neuroprotection for primary rat neurons. However, autophagy and neuroprotection may be either dependent or independent processes, since you have not studied the signaling mechanisms.

Summary.

Current article scientific novelty is very poor. Originality lies in the suggestion autophagy is a key mechanism for neuroprotective effect, but that based on unconfident data and the investigations are insufficient.

Author Response

Question

1. Line 147. How did you control that the astrocytes reached confluency?

In mixed glial cultures, astrocytes exhibit a flat and dark morphology and adhere to the bottom of the culture flask. Confluency is typically assessed morphologically, as the astrocytes uniformly cover the flask surface.

Comments

2. The references list and in-text refs are formatted not according to the journal rules. See Reference List and Citations Style Guide for MDPI Journals on the journal website.

We have re-formatted the references according to the journal rules and number them in the text. Thanks.

3. Figure 4B does not contain column names.

Thanks for spotting this error. We now add on.

4. Line 390, 401. Authors claim that the Lc3b-II expression was significantly lower after 24 h OGD-ACM treatment than that after 6 h OGD-ACM treatment. However, the confidence intervals of the obtained data overlap, that does not allow make this statement for sure. The same for Glut1 and Pfkfb3 upregulation on figure 5.

We appreciate the observation that LC3B-II expression was not significantly different between the 6 h and 24 h OGD-ACM treatment groups in Figure 8B. We have revised the figure and/or text accordingly to reflect this finding

Figure 5 presents qPCR data, where the statistical significance for Glut1 and Pfkfb3 gene expression is shown relative to the normoxia control group. We did not report any statistical comparisons between the 6 h and 24 h OGD treatments for these genes. VEGF is the only gene for which a significant difference was observed between the 6 h and 24 h treatments, and this has been indicated with a hash symbol (#) in the figure.

5. Despite on unconfident Lc3b-II upregulation, p62 protein still downregulated, that witnessed autophagy occurs in OGD astrocytes. Based on that, authors suggest autophagy as a key mechanism in ACM-mediated neuroprotection for primary rat neurons. However, autophagy and neuroprotection may be either dependent or independent processes, since you have not studied the signaling mechanisms.

Although the upregulation of LC3B-II appears inconclusive, the concurrent downregulation of p62 suggests that autophagy is indeed occurring in OGD-treated astrocytes. Based on these findings, we propose that autophagy is a key mechanism underlying the neuroprotective effects of ACM on primary rat neurons. However, it should be noted that autophagy and neuroprotection may represent either interdependent or independent processes, particularly in the absence of direct investigation into the underlying signaling pathways.

Summary

Current article scientific novelty is very poor. Originality lies in the suggestion autophagy is a key mechanism for neuroprotective effect, but that based on unconfident data and the investigations are insufficient.

While the study touches on a potentially interesting aspect of astrocyte-mediated neuroprotection, namely the involvement of autophagy, the current evidence remains preliminary and somewhat inconclusive. The novelty of the work would be strengthened by more robust data and a clearer mechanistic exploration of the proposed link between autophagy and neuroprotection. As it stands, the manuscript raises an important hypothesis but requires further investigation to substantiate its claims.

Reviewer 3 Report

Comments and Suggestions for Authors

The article is devoted to the study of the reaction of primary astrocytes of the rat cerebral cortex to oxygen-glucose deprivation, and the neuroprotective potential of the environment was assessed. It has been shown that astrocytes exhibit an adaptive, time-dependent response to ischemic stress and release soluble factors that can provide neuroprotection.

It is important to note that the work has demonstrated at a qualitative level the therapeutic potential of influencing astrocyte-mediated signaling pathways to increase neuron survival after ischemic stroke.

The article is based on a wide range of sources, its structure is logical, the text is easy to read, and the visual drawings are easy to interpret. In general, the article is written in an accessible and understandable way, and its content will certainly be interesting for readers. 

Author Response

Thank you for your kind and insightful feedback on our manuscript—it is greatly appreciated

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Authors typed some proteins contained in ACM, but did not study them separately to clarify which one or several of them related with the neuroprotective effect after OGD.