Escrever um comentário

Synthesized by: Paula Santos-Otte, Laurent Jutras-Dubé and Anna Oliveras Martinez

In this preprint, Voorbraeck and colleagues use primary neuronal and astrocytic mono- and co-cultures derived from an Alzheimer’s disease (AD) mouse model to investigate the contribution of astrocyte-to-neuron mitochondrial transfer to neuronal bioenergetics, and provide evidence that this mechanism may be impaired in early AD. The authors show that mitochondrial transfer is mediated by mitochondrial extracellular vesicles (mitoEVs) and that it can modulate neuronal bioenergetics. In primary cultures of the AD mouse model, before global bioenergetic decline, neurons display synaptic energy deficits and astrocytes increase mitochondrial network dynamics. Proteomics profiling of WT mitoEVs shows enrichment in proteins supporting oxidative phosphorylation, lipid metabolism, and redox homeostasis, while AD mitoEVs are depleted in proteins supporting bioenergetic health. Finally, the authors prove that mitoEV neuron to astro transfer is compromised in early AD and show that WT mitoEvs have the potential to restore bioenergetic failure in AD neurons. Although neuron-to-astrocyte mitochondrial transfer has been shown to play an important role in neuronal health, this work broadens the repertoire of astrocyte-mediated mechanisms supporting neuronal bioenergetics and identifies a potentially targetable pathway that could inform the development of early interventions for AD.

Overall, we believe that the paper’s strengths outlined below make their findings highly valuable to the scientific community:

• This work adds valuable insights into the basic mechanisms of neuronal bioenergetics regulation, which is especially timely considering the current discussion on how tightly neuronal bioenergetics is regulated in the brain and how its failure might represent the very first step towards early synaptic vulnerability and neurodegeneration.

• Some of the conclusions are well supported by the results. In particular, we find the rescue experiment (Figure 6) to be convincing and informative.

• The preprint highlights that in early AD pathogenesis, prior to AB aggregation and neuronal death, neurons and astrocytes are already suffering from metabolic alterations leading to engagement in compensatory mechanisms that could relieve such metabolic stress. This could have major implications for future therapeutic studies aiming to identify early metabolism-related biomarkers and early disease interventions.

We find that the overall flow of the manuscript is good and that most of the major conclusions are well supported by the data presented. However, there are a few major claims that we believe are not fully supported:

• We recommend toning down the claim “mitochondrial network in AppNL-G-F astrocytes is less interconnected”, since none of the metrics presented in Fig. 2c & e show a statistically significant decrease in connectivity for AppNL-G-F astrocytes.

• The claim “astrocyte-to-neuron mitochondrial transfer occurs at axons” suggests that transfer exclusively happens at the axons, which is not directly addressed by the experiments presented. The experimental design forces the transfer to occur at axons but does not address whether this is the preferential site where transfer occurs. We suggest toning down this claim and reorganizing the figure. We suggest starting with the overall neuron-astrocyte transfer experiments (Fig 3c-j) before showing Fig 3a-b to prove that it can also happen directly at the axonal sites, which would justify the results shown in Fig 1e,f i.e., that ATP is only reduced in synapses and not in the whole axon. Moreover, regarding the data shown in Fig 3b, we consider that to fully support the claim it would be necessary to show more than just one example. We recommend including a quantification of the frequency of the event, for example the number of astrocytic mitochondria per neurite.

• Regarding the section “Neuron-to-astrocyte mitochondrial transfer and fate in astrocytes”, we consider that the major conclusions of this section are not clearly stated. We recommend summarizing them in the section’s last paragraph, and stating the section’s main claims in its title, as in the rest of the results section.

• Some reviewers also find that it is not clearly stated how the APP KI mouse model reflects an early AD model. As this is one of this work’s defining concepts, we recommend that the authors expand on this crucial feature of their model, either in the introduction or with supporting data in the results section.

Minor comments:

• We suggest discussing beta-oxidation and lipid metabolism in the introduction since it is relevant for the results, and it is highlighted in the conclusions, the abstract and the discussion.

• We would refrain from naming the main author of references 15 and 16.

• In the introduction’s last paragraph, the authors could be more specific in their conclusions. For example, to which kind of neuron-astrocyte transfer, and to which AD conditions are the authors referring?

• In the last 2 results sections, we suggest focusing on synthesizing the data and we would recommend moving interpretations to the discussion.

• A description of all mitochondrial features shown in figure 2, including mitochondrial complexity, dynamics and independent components, would be appreciated. We also recommend being more specific when claiming that “AD astrocytes exhibit a less complex but more dynamic mitochondrial network” in the results section and that “astrocytes exhibit a more dynamic but metabolically preserved mitochondrial network” in the discussion.

• In Fig 3h, we suggest including representative examples of fully integrated, attached, and non-associated free structures that match the quantification, similarly to Fig 2b.

• In Fig 4b, the authors should explain why the lack of TOM20 is expected.

• In Fig 4c-e, we suggest moving the interpretation of the results to the discussion and explaining thoroughly how the data presented in this study is “contradicting previous data suggesting that stress and disease conditions enhance EVs secretion”. As shown in Fig 3g, “astrocytes exposed to AppNL-G-F neurons upregulated the expression of genes associated with mitochondrial transport”, suggesting that neuronal signals which flag energetic failure are needed to enhance mitoEV secretion from astrocytes.

• In the PCA plot of Fig 5a, we believe that while there is a separation between the two populations, this separation is not “clear” as stated in the manuscript, since separating the two populations would require drawing a nonlinear curve.

• In Supplementary Fig S5a, b, the rationale behind choosing the appropriate treatment dose could be better stated. As reviewers, we had some questions: Is 12pg selected because it is the highest concentration for which WT neurons do not display significant changes in ATP levels? Why use total intracellular ATP levels as readout, when changes are spatially restricted to pre-synapses? Why does the chosen treatment with WT mitoEVs reduce ATP levels, instead of increasing them?

• We recommend highlighting this work’s novelty in the discussion. For example, the authors show that in early AD pathogenesis, prior to AB aggregation and neuronal death, neurons and astrocytes are already suffering from metabolic alterations, leading to engagement in compensatory mechanisms that could relieve such metabolic stress. This result puts the focus of future studies on metabolism-related biomarkers or targeted therapies.

Comments on reporting and data visualization

• We encourage the authors to provide a rationale for the methodological choices across neuronal and astrocytic experiments. For example, why was qPCR used in astrocytes and WB in neurons, and why was Seahorse analysis performed only in astrocytes?

• Please report the number of experimental replicates (N) and statistical tests used in Figures 3 and 4, and where missing in Figures 5 and 6.

• Please provide the methodological details for the experiments shown in Fig. 1a, Fig. 2j, and Fig. 3c, g.

• In Fig. 1c, we suggest including statistical analyses to support the absence of significant differences in expression. For the right-hand panel, please also include a housekeeping gene and clarify the normalization procedure.

• In Fig. 1d, we noted a potential discrepancy: the main text refers to ATP synthase, whereas the y-axis and Supplementary Fig. S1b refer to MitoDsRed signal. We encourage the authors to clarify and correct this if necessary.

• In Fig. 1e, f, we suggest including representative images illustrating the segmentation of synaptic and total signals.

• For Fig. 3e, we suggest considering a bar plot showing individual measurements, which may provide a clearer representation of the data. Please also report N and the statistical test used.

• For Figures 4 and 5, we suggest using bar plots when fewer than three GO terms are shown, as dot plots with very few data points are difficult to interpret. Where sufficient proteins are detected, volcano plots may provide a more informative visualization. We also suggest considering whether panels 4i, j are necessary, as the main conclusion appears to be supported by the preceding analyses.

• Supplementary Fig. 4e is not referenced in the main text. We also suggest improving its resolution and simplifying the network visualization by focusing on nodes most relevant to the molecular conclusions. A clearer description of the main findings would further facilitate interpretation.

• In Fig. 6i, we noted a potential issue with the reported statistical analysis. Could the authors please confirm whether the difference between WT and APP mitoEV treatment is indeed non-significant and verify the reported statistical test and p-value?

Conflicts of interest of reviewers

• The reviewers declare there are no conflicts of interest.

We wish the authors all the best on continuing this line of research!

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.